Citation
An ecological study of Perognathus Fasciatus

Material Information

Title:
An ecological study of Perognathus Fasciatus
Creator:
Beebe, Chelsea K. ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (58 pages). : ;

Subjects

Subjects / Keywords:
Olive-backed pocket mouse ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The Olive-backed Pocket Mouse (Perognathus fasciatus) is an understudied species whose southern range occupies the Colorado Front Range and eastern plains. According to the limited published literature, this species is not often captured in large numbers. In order to better understand this species conservation status as well as gather ecological base line information on a southern population, we estimated population density, home range area, and ground cover composition within home ranges on a known population of P. fasciatus at the Plains Conservation center south of Strasberg, Colorado. Using distance sampling and a trapping web, we measured population densities at 6.9 individuals /ha in the spring (CV=17.7 percent) and 7.4 individuals/ha in the summer (CV=17. percent). Using radio telemetry to locate nightly movements, home range areas were calculated using kernel density estimations at 95 percent probabilities and ranged from 0.060.84 ha, with an average of 0.395 ha. The ground cover within the 50 percent home range kernels was significantly different between spring and summer with litter and grass and forbs more dominant than bare ground. Policy makers can use this novel base line information to manage this uncommon species and its diminishing habitat.
Thesis:
Thesis (M.S.)--University of Colorado Denver. Biology
Bibliography:
Includes bibliographic references.
System Details:
System requirements: Adobe Reader.
General Note:
Department of Integrative Biology
Statement of Responsibility:
by Chelsea K. Beebe.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
898221644 ( OCLC )
ocn898221644

Downloads

This item has the following downloads:


Full Text
AN ECOLOGICAL STUDY OF PEROGNATHUS FASCIATUS
by
CHELSEA K. BEEBE
B.A., University of Colorado, Boulder, 2004
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Biology
2014


This thesis for the Master of Science degree by
Chelsea K. Beebe
has been approved for the
Department of Integrative Biology
by
Michael B. Wunder, Chair
Laurel M. Hartley
Cheri Jones
Lance Carpenter


Beebe, Chelsea, K. (M.S., Biology)
An Ecological Study of Perognathus fasciatus
Thesis directed by Professor Michael B. Wunder
ABSTRACT
The Olive-backed Pocket Mouse [Perognathus fasciatus) is an understudied
species whose southern range occupies the Colorado Front Range and eastern
plains. According to the limited published literature, this species is not often
captured in large numbers. In order to better understand this species conservation
status as well as gather ecological base line information on a southern population,
we estimated population density, home range area, and ground cover composition
within home ranges on a known population of P. fasciatus at the Plains Conservation
center south of Strasberg, Colorado. Using distance sampling and a trapping web,
we measured population densities at 6.9 individuals /ha in the spring (CV=17.7%) and
7.4 individuals/ha in the summer (CV=17.8%). Using radio telemetry to locate nightly
movements, home range areas were calculated using kernel density estimations at 95%
probabilities and ranged from 0.06-0.84 ha, with an average of 0.395 ha. The ground
cover within the 50% home range kernels was significantly different between spring and
summer with litter and grass and forbs more dominant than bare ground. Policy makers
can use this novel base line information to manage this uncommon species and its
diminishing habitat.
m


The form and content of this abstract are approved. I recommend its publication.
Approved: Michael B. Wunder
IV


ACKNOWLEDGMENTS
I would like to thank my advisor Mike Wunder for all his teachings and
support. He helped me understand statistics, a feat I never thought I could manage
and allowed me to make my own mistakes and learn from them. I would also like to
extend a special thank you to Lance Carpenter, the most hard core Biologist I know,
and by extension his colleges at Colorado Parks and Wildlife, for hiring me as a
technician and creating the original project idea. Thanks to the Plains Conservation
Center, especially Tudi Arneill, for permitting me to work on their site and
graciously allowing me to live rent-free on site throughout the trapping/tracking
efforts. I cannot express my gratitude to my friends and colleagues, too many to
name, that volunteered to help me in the field and weathered all nighters, rattle
snakes, and gusting prairie winds. I would also like to thank Laurel Hartley and
Cheri Jones for being on my committee and offering their feedback and ideas
throughout the entire process of this project. Further acknowledgment to my
friends and colleagues in the Department of Integrative Biology and the Ecology and
Evolutionary Biology group for collaboration, encouragement, and countless hours
at the coffee shops of Denver, especially Libby Pansing, Mari Majack, Jill Pyatt,
Rebecca Bryan, Andrew Doll, Sarah Blakeslee, and Aaron Wagner. And lastly to my
family, who helped me in the field and never ceased to offer love and support.
v


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION
Background ..................................................1
Home Range...................................................4
Rarity.......................................................4
Small Mammals in a Changing Landscape: The Case
of the Prebles Meadow Jumping Mouse.........................9
Study Objectives............................................11
II. METHODS........................................................12
Study Site Description......................................12
Study Design................................................12
Radio Telemetry.............................................14
Vegetation Survey...........................................16
Density Analysis............................................17
Home Range Analysis.........................................17
Tables and Figures .........................................19
III. RESULTS........................................................22
Density and Abundance of P.fasciatus........................22
Home Range Size.............................................23
Vegetation..................................................23
Tables and Figures .........................................24
IV. DISCUSSION.....................................................28
vi


Density and Abundance....................................28
Home Range...............................................30
Ground Cover.............................................31
Dominant Plant Species...................................32
Suitable soils for P.fasciatus in Colorado...............33
Rare Status Assessment...................................34
Conservation of Rare Species in Diminishing Habitat......36
Tables and Figures.......................................37
REFERENCES..................................................40
APPENDIX....................................................45
vii


LIST OF TABLES
Table
111.1. Total captures and recaptures by species..............................24
111.2. Density and abundance of Perognathus fasciatus by season..............24
111.3. AIC values for model selection for Perognathus fasciatus..............24
111.4. Home range areas of individual animals................................25
111.5. Home range estimations of other Perognathus sp........................25
111.6. Medium and interquartile range of ground cover........................25
IV. 1. Density and abundance of Reithrodontomys megalotis....................39
Al. Density of all rodents at the PCC by season..............................45
A2. AIC values for model selection for all rodents by season.................45
A3. Reproductive status of P. fasciatus in 2012..............................45
A4. Age of P. fasciatus in 2012..............................................46
A5. Species Richness between 2010 and 2012...................................46
A6. Comparison of ground cover between 2010 and 2012 trapping efforts........46
A7. Soil types returned in the OB PM area of occurrence analysis.............47
viii


LIST OF FIGURES
Figure
II. 1. Research study area...................................................19
11.2. Trapping web...........................................................20
11.3. Transmitter and harness................................................20
11.4. Vegetation sampling design.............................................21
11.5. Vegetation grid design.................................................21
111.1. Kernel density polygons at 95% on base map...........................26
111.2. Bare ground, grass/forb, and litter boxplots.........................27
IV. 1. Aerial view of study site with trapping web and capture area.........37
IV.2. Kernel density polygons at 50% with overlap...........................38
IV.3. Suitable soils for P.fasciatus in Colorado............................39
Al. 95% kernel polygons for all animals by season...........................48
A2. PCC study site including the 2010 transects.............................49
IX


LIST OF ABBREVIATIONS
AIC Akaike information criterion
AO Area of occurrence
CGAP Colorado Gap Analysis Program
CNHP Colorado Natural Heritage Program
CSURF Colorado State University Research Foundation
EO Extent of occurrence
GPS Global positioning system
IACUC Institutional Animal Care and Use Committee
IUCN International Union for Conservation of Nature
LSCV Least squared cross validation
NRCS Natural Resource Conservation Service
OBPM Olive-backed Pocket Mouse
PIT Passive Integrative Receiver
PCC Plains Conservation Center
PMJM Prebles Meadow Jumping Mouse
USD A United States Department of Agriculture
UTM Universal Transverse Mercator
x


CHAPTER I
INTRODUCTION
Background
The Olive-backed Pocket Mouse, OBPM, (Perognathus fasciatus) is an
understudied species that occurs on the short/mixed grass prairies of the Colorado Front
Range. This uncommon species has not been the subject of many demographic studies
presumably because of the challenge of finding and capturing enough individuals for a
suitable sample size. The geographic distribution range for this species extends from
southern Canada south through the Dakotas to the eastern plains of southern Colorado
(Manning and Jones, 1988). Historical captures in Colorado have been reported from
numerous sites along the Front Range of the Rocky Mountains in Larimer, Jefferson,
Fremont, and Custer counties in Colorado.
The family Heteromyidae includes several species of pocket mouse. Whereas
there is limited literature about P. fasciatus, studies on other heteromyids can be used for
a better understanding of the genus Perognathus. Pocket mice are thought to be
primarily granivorous, collecting seeds in their cheek pouches and caching either
underground or on the soil surface (Randall, 1993). Perognathus fasciatus eats various
grass seeds and other herbaceous plants (Manning and Jones, 1988). Pocket mice tend to
become dormant in the colder winter months, retreating to underground burrows and
going into torpor, rousing periodically to feed on food stores (Wrigley et al., 1991). After
a period of dormancy, pocket mice emerge in the early spring to reproduce before the dry
months of summer, but can produce additional litters if favorable conditions exist
(Randall, 1993; Wrigley et al., 1991). Gestation time is approximately four weeks with
2


2-6 offspring in the average litter (Manning and Jones, 1988). Some heteromyids enter
estivation, a state similar to torpor initiated by very hot temperatures (Merritt, 2010, p
158). The survival and life span of P.fasciatus is not known, but Hayden and Lindberg
(1976) estimated life span of P. longimembris, a similar sized pocket mouse, to be
approximately one year in the wild, with some instances of longevity extending up to
three years. As residents of desert ecosystems, much attention has been afforded to the
Heteromyidae as a model of coexistence and sympatric speciation in harsh environments.
Randall (1993) synthesized foraging strategies, microhabitat and space use, and general
courtship behavior. Neiswenter and Riddle (2001) discussed the family evolution
throughout the glacial periods of North America that led to the distinction of P.fasciatus
as a unique species.
Whereas there is some general life history information available for P. fasciatus,
little quantitative information exists on the demography of this species. Population
densities have been estimated as 0.4-4.0 individuals per ha at various northern locations
within North Dakota and Montana (Pefaur and Hoffman, 1974; Genoways and Jones,
1972). No population density study has been conducted in the eastern plains of Colorado.
In populations of the Great Basin Pocket mouse, P. parvus, abundance estimates suggest
a seasonal abundance peak in late summer, with the lowest numbers in late winter early
spring (Jan.-Apr.) (Hedlund and Rogers, 1980).
Estimates of small mammal abundance and density are important for
understanding population dynamics. Estimating wildlife population sizes is critical for
assessing ecosystem health at the community and population levels. In species that are
3


understudied, density can also provide insight into demographic parameters, community
interactions and behavior, as well as home range.
Home Range
W. H. Burt (1943) defined home range as the area traversed by the individual in
its normal activities of food gathering, mating and caring for young. This classic
definition has under gone some modifications over the years by several groups (Fieberg
and Kochanny, 2005). Home range size and selection are affected by resource
availability, interspecific and intraspecific competition. These factors can be influenced
by population density and abundance (Orland and Kelt, 2007; Kotler and Brown, 1988).
The relationships between home range size, extent of overlap, and seasonal shifts are
complex. Habitat type, weather and seasonality, predators-prey dynamics, sympatric
species, and unique species populations all likely play influential roles (Quirici et al.
2010). There is very little literature about P. fasciatus home range size. The greatest
distance between trap locations for successive captures of individuals include 26.5 meters
and 65.7 meters (Manning and Jones, 1988; Pefaur and Hoffman, 1975). Home range
estimates from similar sized Perognathus range from 0.02 0.63 ha (Monk and Jones,
1996Best and Skupski, 1994; Hedlund and Rogers, 1980). Further research on the home
range size of P. fasciatus in the prairies of Colorado is needed because there are no
formal estimates of home range size.
Rarity
Rare species are at a higher risk of extinction than are more common species
(Drever et al., 2012). Species populations that are characterized by low abundances are
more susceptible to demographic stochasticity, genetic inbreeding effects and extreme
4


climate events from which population recovery is impossible. Rare species also pose a
challenge for collecting basic ecological and demographic data. If a species is not of any
particular conservation concern, there might not be any interest in concentrated efforts to
conduct studies to fill knowledge gaps. As a result, rare species are often understudied
with limited accounts of sighting, collection, and tracking. A comprehensive
understanding of all members of an ecosystem can help define phylogenetic relationships
between similar species or reveal mutualistic or competitive interactions that influence
the entire food web.
The OBPM is not often captured in small mammal studies and when it is, it is not
captured in large numbers. The perceived rarity of this species suggests a need for
further investigation. The concept of rarity is deceptively complex, but a basic definition
is a species population with low abundance or limited geographic range (Drever et al.,
2012; Gaston, 1991). Rabinowitz (1981) first described types of rareness in three areas:
small geographic range, narrow habitat specificity, and small/non-dominant population
size. If a species meets any one of these criteria, it could be defined as a rare species.
Variations of these criteria are further used to identify species that are threatened or
endangered. The International Union for Conservation of Nature (IUCN) red list is
widely regarded as the authority on threats to existence of species. Threat level is
assessed by several factors that are similar to the Rabinowitz (1981) definition. The
IUCN examines rate of population decline over time (from historic numbers to current
population estimates), range area (including extent of occurrence and area of occurrence),
population size and fluctuations, and quantitative analysis that considers demographic
5


patterns into the future at established time intervals
fhttp://www.iucnredlist.org/details/42608/0).
Perognathus fasciatus has been described as inhabiting a large geographic range
with a wide range of habitat specificity. It is considered small and non-dominant in its
population size. This form of rarity is consistently sparse over a large range and in
several habitats (Rabinowitz, 1981, p 208). The geographical range has been determined
from captures and museum specimens around North America, likely by a combination of
the range wide occurrence method and the marginal occurrence method (Gaston and
Fuller, 2009). These methods are prone to overestimation of total range size. The extent
of occurrence (EO) and the area of occurrence (AO) dissect the geographical range into
two arenas. The EO is the estimated geographic distribution that uses actual captures as
well as estimations of where the species would be expected based on habitat suitability.
The AO is the actual sites of captures or sightings of the species, further described by
individual home range size and landscape composition (i.e. vegetation, soil type,
topography, climate). The habitat that is used or occupied is a superior representation of
the species true distribution (Gaston, 1991). While the EO of P. fasciatus is well defined,
albeit outdated, the factors used for habitat suitability are less well understood. There are
descriptions of habitat type, but no repeated attempts to quantify ground cover type,
relative floral species abundance, or other aspects used for predicting species presence in
an area. The habitats reported for P. fasciatus include transitional upland grassy
habitats, grassy hillsides among rocky outcroppings, and on [a] sandy floodplain
within 100 yards of the river (Manning and Jones, 1988). Species mentioned include
Populus sp., Bromus sp., Poa sp., Rosa multiflora, Artemisia frigida at one site in North
6


Dakota and Stipa, Bouteloua, Carex at another; Yucca glauca, and Pinus ponderosa in
Nebraska; Artemisia tridentata, Atriplex sp., and Tetrademia sp. in Utah; and Carex,
Bouteloua, and Agropyron in Montana (Manning and Jones, 1988). These descriptions
suggest general prairie landscapes, but because P.fasciatus is not captured in many
trapping studies, there are likely overlooked aspects of the habitat type that are most
important for predicting population existence. According to the IUCN, P.fasciatus is
listed as Least Concern because it is widespread, its populations are secure, there are no
major threats, and it occurs in many protected areas throughout its range
(http://www.iucnredlist.org/details/42608/0T The literature cited in the IUCN description
of the OBPM is primarily from studies published in the 1970s.
Another important consideration when estimating rarity is the probability (p) of
detecting an animal given that it is present in the area being surveyed. In mark-recapture
studies that assume a closed population, a common equation used to estimate abundance
is N=n/p, where N is the population size, n is the observed count of unique individuals,
and p is the probability of detecting an animal given that it is available for capture or
observation. Probability of detection, p, can be estimated by repeated measures of
observation (such as from a mark recapture type of study), or by modeling it as a function
of some covariates, such as distance from observer, intensity of effort, etc. Rare species
that exist at low population densities or that are locally abundant but regionally rare are
difficult to encounter. The OBPM tendency towards inactivity in periods of extreme
temperatures, either by entering torpor or estivation, poses additional challenges toward
population/individual detection. Collecting enough data points for a mark recapture
study can therefore be a challenge (Mackenzie et al., 2005). This reduces the precision of
7


the estimates for p. Distance sampling is an approach that models detection probability as
a function of distance from observer and provides a way to adjust for detection-based bias
in models that are used to estimate population density or abundance. For a small,
nocturnal rodent that is not easy to observe, distance sampling can be conducted with
trapping webs to account for imperfect detection in the estimation of densities in sparse
populations when four assumptions are met (Parmenter et al., 2003). The first
assumption is that all animals at the web center are captured, and that animal movement
is constant throughout the web with no net emigration or immigration. The second
assumption is that all animal captures are individual events. The third assumption is that
individuals are recorded at their original location and finally, that all distances between
traps are measured accurately (Anderson et ah, 1983).
Measurements of population abundance and geographic range allow for
quantitative analysis of rarity, and also provide specific insight for the type of rarity
exhibited (Drever et ah, 2012; Rabinowitz, 1981). Distinctions can be made between
high/low abundance in one region, or high/low abundance across the entire geographic
range and thus inform the subsequent conservation strategies in specific areas, or more
importantly, specific states or counties. The organization Nature Serve classifies P.
fasciatus as secure, however in Colorado it is classified as vulnerable
(http://natureserve.org/getData/index.isp). This specification to the Colorado population
is an example of how jurisdictional boundaries can influence conservation priorities. The
designation of states and countries within the species geographic range can have harmful
and long lasting effects on the survival of the species, as conservation and management
priorities vary across boundary lines. Increased threats at the species range ends or in the
8


center might cause fragmentation leading to reduced gene flow or loss of genetically
diverse edge populations (Drever et al., 2012). Edge populations may carry diverse
alleles, making them more suited to variable conditions that differ from those in the
central part of the range. This genetic diversity can provide valuable traits for the entire
population in shifting geographic ranges that result from climate change (Hutchinson and
Hunter, 1994).
The populations of P.fasciatus with the jurisdictional boundary of Colorado are a
unique subset of the species. Living at the southern edge of the entire geographic range,
this collection of populations might be adapted to the hotter and drier conditions typical
of lower latitudes. This area of the Great Plains is also experiencing higher human
population growth as compared to the more northern states of Wyoming, North Dakota,
and Montana. Overall, the Great Plains of North America are in a state of decline
(Samson and Knopf, 1994). A consequence of rapid conversion of a threatened habitat is
the loss of rare species. Compounding this problem are the threats to rare species at the
periphery of their ranges. Past conservation issues have shown that more baseline
information about the rodent species found in Colorado is needed in order to develop
informed land use plans that consider both human land use and habitat needs for the
native fauna.
Small Mammals in a Changing Landscape:
The Case of the Prebles Meadow Jumping Mouse
In the 1990s, the Prebles Meadow Jumping mouse, PMJM, (Zapus hudsonius
preble) became a species of concern within its limited Colorado and Wyoming range.
This mouse lives in riparian habitat, a habitat type that is relatively rare and at risk from
9


development and pollution along the Front Range of the Rockies and along the western
edge of the Great Plains. Zapus hudsonius preblei is a subspecies of Zapus hudsonius, a
species with a much larger geographic range that extends east across the northern US and
through much of Canada. There were two conflicting research groups that debated the
validity of a subspecies classification, one arguing that Z.h. preblei was a unique
subspecies and the other arguing that it was not, based off independent molecular
genetics studies and taxonomic interpretations (Crifasi, 2007; King et al., 2006; Ramey et
al., 2005).
Proponents of the unique subspecies pushed for quick moving legislation to
protect critical habitat, stressing that the risk of extinction without immediate action was
high. Legislation was passed and the mouse was added to the endangered species list as a
threatened species in 1998 causing a cascade of changes for landowners and developers.
There has since been evidence suggesting that the subspecies Z. h. preblei is not a unique
taxon, but shares gene flow with other subspecies of Z. hudsonius (Malaney and Cook,
2013). At present, the subspecies is recognized and federally protected, but remains
controversial. This debate resulted in copious spending, heightened emotions, and
disruption of development (Crifasi, 2007). Regardless of attitudes on the Endangered
Species Act and its impacts on conservation and development, there is a lesson that can
be related to other small mammals in the Colorado Front Range.
Sound science-based policy demands current and accurate information on the life
histories, ecology, and evolutionary relationships among species within the jurisdiction
administered by policy makers. Had intensive morphometric, ecological, and genetic
studies been preformed on Z. h. preblei prior to over development and degradation of its
10


habitat, more informed decisions regarding listing could have been made in a timely
manner. Conducting post-hoc analysis to answer a scientific question, rather than
hypothesis testing, may lead to controversial interpretations. The Olive-backed Pocket
Mouse also resides in a habitat type that is at risk of increased conversion from human
activities. Should a conflict similar to the PMJMs occur, a body of information on this
understudied species must be made available to all stakeholders. Quantitative and
accurate estimates of vital demographic parameters and home range area can be added to
existing genetic and speciation information and aid policy makers and interest groups in
making informed decisions.
Study Objectives
The objective of this study was to gather basic field data from a known population
of P.fasciatus to measure population density, to calculate home range size, and to
describe ground cover characteristics within estimated home ranges. This study will
contribute to the sparse literature on this species and aid in future prairie management
plans and conservation decisions.
11


CHAPTER II
METHODS
Study Site Description
This study was conducted from 2012-2013 at the Plains Conservation Center
(PCC) West Bijou Creek site near the boundary between Arapahoe and Elbert counties,
near Strasburg, Colorado (39 35' 12.12", -104 16' 38.59). This historically private
rangeland was purchased by the PCC in 2006 (Figure II. 1). The site continues to support
occasional cattle and bison grazing. The vegetation consists primarily of native grasses
(e.g. Pascopyrum smithii) and non-native grasses (Panicum virgatum and Sporobolus
cryptandrus), with a sparsely scattered canopy of rabbitbrush (Chrysothamnus nauseous).
This site was selected because previous survey efforts discovered a population of P.
fasciatus here. This area of the Great Plains receives about 39 cm. of rain annually and
experiences winter low temperatures of -11C and summer high temperatures of 32C.
Study Design
To calculate animal density through distance sampling, we deployed a trapping
web consisting of 16 lines radiating from a center point (distance or d=0). Each line
was 75 meters long with 15 Sherman live traps (3x3.5x9; H. B. Sherman Traps, Inc.,
Tallahassee, Florida) spaced five meters apart. There were another six traps at d=0 for a
total of 246 traps (Figure II.2). We used a measuring tape to ensure accurate five meter
spacing from the center and between subsequent traps. We used a compass to set the
direction of the 16 lines. The center of the web was placed purposefully at a coordinate
where a high density of P. fasciatus was located in 2010 (Colorado Parks and Wildlife,
unpublished). This placement was not random because the species is rare; we focused
12


our efforts in an area where a known population existed. Due to limited equipment and
personnel, we placed just one web in the same location for each seasonal effort.
We conducted trapping in each of the four recognized seasons, spring (5-17 May
2012), summer (1-10 August 2012), autumn (2-11 November 2012), and winter (2-12
January 2013). Trapping events consisted of two sets of four consecutive trap nights,
with a two-night break between. All traps received a small ball of Polyfill for insulation
and were baited approximately 1 hour before dusk with moistened birdseed (Kaytee
brand wild birdseed). Traps baited with moistened birdseed captured more Heteromyids
than rolled oats and peanut butter in our 2010 trapping effort (CPW, 2010 unpublished).
We checked the traps at dawn each morning. During most of the autumn session and all
of the winter session, traps were also checked between 00:00-01:00. At this time all
animals were processed, released, and the traps reopened and then checked again at dawn
to minimize exposure. Individuals that were recaptured in a single night were
immediately released and only the data collected from the first capture were recorded.
All captured animals were identified to species prior to being weighed (to the
nearest O.lg). We measured and recorded standard morphometries including tail length,
left foot length, ear length and noted the reproductive status for each animal. All
captured animals were uniquely marked. Individuals of Perognathus sp., Chaetodipus
hispidus, Microtus ochrogaster, and Dipodomys ordii were marked with a Passive
Integrative Receiver (PIT) tag (Biomark HPT 8, 8.4 MM 134.2 kHz FDXB) that was
inserted subcutaneously between the shoulder blades. Animals were anesthetized with an
isoflurane inhalant prior to insertion; approximately 3mLs of isoflurane was applied to a
cotton ball that was placed inside a metal tea ball. The animal and the tea ball were
13


secured in a sealed Tupperware container with punctured holes, and we visually
monitored each animal for signs that the anesthesia had taken effect (e.g. until it was
unable to maintain an upright posture). Once anesthetized, the animal was removed from
the container and we inserted the PIT tag using a hollow-tipped needle (Biomark N165
needle for MK165 implanter). Because we had a limited number of PIT tags, individuals
of the most abundant species, Reithrodontomys sp., were marked by a unique
combination of toe amputations. No more than one toe was taken from the front and
back, and we did not remove medial fore fingers. All animal protocols were approved by
the University of Colorado Denver IACUC (protocol# 101836) and were in accordance
with the guidelines established by the American Society of Mammalogists for the use of
wild mammals in research (1998).
Radio Telemetry
Perognathus fasciatus individuals were captured as part of a trapping effort that
was independent from the trapping web effort to estimate density and abundance that was
described above. We placed traps in areas where individuals had been previously
captured for the mark-recapture study, supplied with Polyfill bedding, and baited with
moistened birdseed, and opened approximately one hour before dusk. Traps were
checked at dawn.
Radio transmitters weighing 0.4 g (model BD-2N; Holohil Systems Ltd., Carp,
Ontario, Canada) were attached to P. fasciatus using a harness that was constructed of
0.5mm elastic cord (Bead landing) threaded through a 0.3 mm beading crimp tube
(Beadalon) into a figure eight. The transmitters were then attached to the beading
crimp tube using Gorilla Glue brand adhesive and the harness looped over the fore limbs
14


with the transmitters resting on the animals dorsal side, as one would wear a backpack
(Figure II.3). Each animal was placed in a small plastic terrarium to observe movement,
ensuring a proper harness fit and full range of motion. All animals were released at the
point of capture once fitted with the transmitter backpack.
Animals were subsequently relocated at hourly intervals for the 7-9 hours
between dawn and dusk. We used a receiver and a yagi antenna to locate the animals and
approached the animal within five meters. This distance was initially calibrated using a
placed transmitter at a five meter distance and observing the signal strength as a measure
of gain and volume on the receiver. We marked the spot with a pin flag, recorded the
direction of the strongest signal with a compass bearing, and then quickly and quietly
moved approximately ten meters to a location parallel to the signal and repeated the
measurement. We then returned during the day and recorded the point at which the
bearings intersected using a measuring tape and a GPS (Garmin, GPSmap 60). We also
placed a color-coded pin flag that corresponded to the animal at the intersection point and
removed the other two triangulation pin flags. This allowed for a clear visualization of
each animals movement locations throughout the site. We also located each animal
underground during the day.
Radio telemetry was conducted in the spring and summer only, as no P. fasciatus
were captured in the fall or winter. In the spring, between 24-30 May 2012, three
individuals were harnessed. We collected data on two and lost one individual to a
predator before any data were collected. In the summer, on 17 August 2012, 4 individuals
were harnessed but all transmitters fell off animals after 3-4 days of data collection. The
procedure was repeated with new transmitters on 11 September 2012, on 4 animals, three
15


of which were the same animals from the first summer trial, and one new capture. The
transmitters fell off these animals again after 2-3 days.
Vegetation Survey
Directly following each radio telemetry seasonal effort, we collected vegetation
data from the area where the animals were relocated to characterize the habitat used. We
used a 0.5m x lm Daubenmire frame and estimated percent composition of bare ground
(ground that contained no visible plant material), grasses and forbs, litter (dead plant
material), and woody shrubs. We also recorded the height of the tallest shrub within the
frame and the width of the shrub at the widest point. Where shrubs were present, we also
estimated both the percentage coverage of the shrubs and the undergrowth, so total
composition exceeded 100% in some plots. These categories were selected because they
are large-scale indicators that can be quickly assessed during trap site selection. Previous
research also indicates that bare ground, shrubs, and litter are important characteristics for
Heteromyid foraging and locomotion (Randall, 1993).
The ground cover survey area was created around the recorded triangulation
points of nightly animal movement. We used a visual assessment of the color-coded pin
flags to select the survey areas. Since there was much clustering of location points, we
delineated a rectangle around obvious clusters, leaving at least a two meter border
between the outer flags and the survey edge. We then divided the rectangle into two
meter by two meter squares with a measuring tape, and sampled each square of the large
rectangle with the Daubenmire frame (Figure II.4). Each surveyed square received an
identifying code. Points that did not easily fit into the survey rectangles were treated as
outliers. Using the outlying triangulation point as the center, we created a three by three
16


square survey around the point and repeated the Daubenmire sampling procedure. The
outer four corners of all rectangle plots were recorded with a GPS. Coordinate points of
each Daubenmire frame were not recorded because the error on the GPS was larger then
the size of the sampled square.
Using the outer four comers UTMs of the rectangle plots, we assigned GPS
coordinates to each sampled square in ARC Map 10.0 with the fishnet tool. The output
fishnet grid points were assigned to the corresponding vegetation sample frame based on
the identifying code assigned in the field (Figure II. 5).
Density Analysis
We used Program DISTANCE version 6.0 to calculate population density. We
evaluated the fit for a global detection function for all P. fasciatus and for each unique
season using uniform, half normal, hazard-rate, and negative exponential key functions
with a cosine, simple polynomial, and hermite polynomial series expansions. The most
parsimonious model was selected using Akaike Information Criterion (AIC) values
(Burnham and Anderson 2002). We performed the analysis using both the full set of
captures and the data truncated, with all captures past 65 meters excluded. The exclusion
of encounters from the outermost rings adjusts for movement of individuals into the web.
Home Range Analysis
For each animal, we estimated a kernel home range utilization distribution at 95%
and 50% probabilities using the Animal Movement extension of ARC View 3.2 (P. N.
Hooge and B. Eichenlaub. 1997. Animal movement extension to arcview. ver. 1.1.
Alaska Science Center Biological Science Office, U.S. Geological Survey, Anchorage,
AK, USA.). These values were used because this is an unofficial standard throughout
17


home range literature and will provide results that are comparable between studies.
Additionally, 95% excludes the outer 5% of movement points, points that may indicate
an uncommon foray from the home range core. The smoothing parameter (H) was
calculated automatically using the Least squared cross validation (LSCV) method.
Because we were interested in characterizing areas of high use, we used vegetation plots
that were sampled within the estimated 50% kernel for ground cover analysis. Using
ARC Map 10.1, we extracted the vegetation point data that fell within the boundaries of
the 50% kernel with the Extract point tool. The median and interquartile ranges of each
percent ground cover category were calculated using Microsoft Excel for each animal.
18


Tables and Figures
19


Figure II.2: Trapping web consists of 16 lines labeled A-P that radiate from a center
point. Each spoke has 15 Sherman Live Traps spaced five meters apart beginning at
d=5m and ending at d=75m. There are six traps in the center at d=0m.
20


0.5m
Figure II.4: Representation of vegetation survey rectangle, composed of 2m x 2m
squares, with Daubenmire frames (0.5m x lm rectangles within). The red flags indicate
movement observation points of P.fasciatus individuals.
Spring
Triangulation Points
Individual
D
W
Summer
Triangulation Points J
Individual
A
A B
E
F
Figure II.5: Vegetation sampling grids created in ArcMap using the fishnet tool, with
animal observation points overlaid.
21


CHAPTER III
RESULTS
In spring, we recorded 156 captures of 60 unique individuals of the following
species: Perognathus fasciatus, Peromyscus maniculatus (Deer mouse), Reithrodontomys
megalotis (Western Harvest mouse), R. montanus (Plains Harvest mouse), Microtus
ochrogaster (Prairie vole), Neotoma mexicanus (Mexican woodrat), Chaetodipus
hispidus (Hispid pocket mouse), and Dipodomys ordii (Ords kangaroo rat). In the
summer we recorded 79 captures of 25 individuals including P. fasciatus, P. maniculatus,
R. megalotis, R. montanus, M. ochrogaster, and C. hispidus. In the autumn we recorded
185 captures of 68 individuals including R. megalotis, R. montanus P. maniculatus, and
D. ordii. In the winter we recorded 151 captures of 76 individuals including R.
megalotis, P. maniculatus, and/), ordii (Table III. 1).
Density and Abundance of P. fasciatus
The density and the abundance of P. fasciatus were similar in the spring and the
summer (Table III.2). A uniform key function model with a polynomial expansion was
selected as the most parsimonious model for both seasons based on the lowest value for
AIC among competing models (Table III.3). The data were not truncated and analysis
was based on exact measurements. The shape of the function was constrained as weakly
monotonically non-increasing. The model predicted an estimated density of 6.9
individuals/ha in the spring (CV=17.7%) and 7.4 individuals/ha in the summer
(CV=17.8%). The detection probability was modeled and estimated by stratum, or
season, and returned values of 0.308 for the spring and 0.257 for the summer.
22


Home Range Size
We collected home range data on six P. fasciatus individuals, two in the spring
and four in the summer. The home range areas were similar between seasons (Table
III.4), ranging between 0.06-0.84 ha, with an average of 0.395 ha. There was much
overlap of the 95% kernels for the animals in both seasons (Figure III. 1), but less so at
the 50% estimation. All male home range areas (mean 0.66 ha, N=3) were larger than
those for the females (mean 0.13 ha, N=3).
Vegetation
We found the median of the measured ground cover percentages for each animals
home range (Figure III.2). We also calculate the interquartile range for each animal
(Table III.6). Shrub cover was minimal and was excluded from the analysis. The five
most dominant species in both seasons were Sporobolus cryptandrus (Torr.) A. Gray
(Sand dropseed), Heterotheca canescens (DC.) Shinners (grey golden-aster), Pascopyrum
smithii (Rydb.) A. Love (Western wheatgrass), Aristidapurpurea Nutt. (Purple
threeawn), and Bromus sp. (Cheat grass).
23


Tables and Figures
Table IEL1: Summary of unique individuals (I) captured per season and total recaptured
(R) animals. _______________________________________________
Spring Summer Fall Winter
Species I R I R I R I R
C. hispidus 1 4 1 6 0 0 0 0
D. ordii 2 3 0 0 2 1 1 1
M. ochrogaster 4 2 1 1 0 0 0 0
N. mexicanus 1 0 0 0 0 0 0 0
P. fasciatus 8 26 7 20 0 0 0 0
P. manicidatus 3 5 1 2 4 3 5 5
R. megalotis 41 93 16 41 59 91 67 104
R. montanus 4 4 2 1 2 2 0 0
Table III.2: Density (individuals per hectare) and abundance for P.fasciatus in spring
and summer.
Season D %CV df 95% Confidence interval
Spring 6.9355 17.8 38.51 4.8516 9.9146
Summer 7.4131 17.7 30.47 5.1925 10.635
N
Spring 12 17.8 38.51 9 18
Summer 13 17.7 30.47 9 19
Table III.3: Program DISTANCE models used to estimate density of P.fasciatus. The
most parsimonious model has the lowest AIC score.
Model No. of parameters AAIC AIC score
Uniform simple polynomial 3 0 490.81
Negative cosine 4 6.06 496.87
Half normal simple polynomial 2 7.27 498.08
Negative simple polynomial 2 7.75 498.56
24


Table III.4: Home range area by animal as calculated by the animal movement extension
in ARC 3.2 at the 95% kernel.
P.fasciatus Individual Season Sex 95% fixed-kernel (ha)
D Spring Male 0.39
W Spring Male 0.75
A Summer Female 0.16
B Summer Male 0.84
E Summer Female 0.06
F Summer Female 0.17
Table III.5: Home range areas of other Perognathus species that show some portion
of their geographic range in Colorado (Studies that measured home range size were
not conducted on Colorado populations').__________________________________________________
Colorado Perognathus sp. Home Range (Hectares)
Perognathus fasciatus .06- .84
Perognathus flavescensa .02- .05
Perognathus flavusb .11 .63
Perognathus parvusc .07- .37
a Monk and Jones 1996 b Best and Skupski 1994 c Hedlund and Rogers 1980
Table III.6: Median and interquartile range for all ground cover type by individuals
tracked. Results are from ground cover within the 50% kernels._______________________
Spring Summer
INDIVIDUAL D W A B E F
P=H £ Grass/Forbs Median 35 45 45 35 30 30
> o Interquartile Range 25 20 15 25 25 20
o Litter Median 45 40 45 47.5 55 60
o o Interquartile Range 25 20 35 25 22.5 20
Bare ground Median 15 10 5 5 10 0
Interquartile Range 20 20 10 15 15 3.75
25


0 12.5 25
T=---~
, k"
Meters!
50 !
95% Kernels
| | D (male)
W (male)
|__A (female)
E (female)
F (female)
N
A
-1 w. \
>;'Vr,.
Lz____-__________<____r_________________L
Source: Esri, DigitalGSIobe, GeaEya, i^ibad, USDA, USGS, AEX,
Getmapplng, Aerogrld.flGN, IGP, swlsstopo, arid the GfShJser *
Community * *
Figure IIE.l: 95% kernel density polygon for all individuals in both spring and summer,
with imagery base map of landscape.
26


Percent Cover Percent Cover Percent Cover
Bareground Cover
Animal
Grass/Forb Cover
O
00
O
CD
O
"3-
O
CM
Animal
Litter Cover
o
o
o
00
o
CD
O
o
Animal
Figure III.2: Median ground cover with first and third quartiles, including bare ground,
grass/forbs, and litter; by animal for all animals tracked.
27


CHAPTER IV
DISCUSSION
Density and Abundance
To our knowledge, this study provided the first estimation of P.fasciatus
population density in Colorado. Knowledge of this basic population parameter can be
used to estimate actual species numbers that provide valuable information for
conservation and management strategies involving habitat preservation and space
allocation. The estimated density was higher in this study than in previous studies. The
differences in densities might reflect different methods used for estimation. Our study
used distance sampling, whereas previous studies calculated density by dividing
individuals caught by total trapping area (Pefaur and Hoffman 1974, Genoways and
Jones 1972). When the data from this study were treated in this way, density estimates
were lower (spring was 4.53/Ha and summer 3.40/Ha), falling close to the upper end of
the previous estimates. Early estimation methods that did not consider the probability of
detection and required estimates of the effective trapping area are biased towards
overestimation (Anderson et al., 1983). Another benefit of distance sampling for the
estimation of population density is the calculation of a 95% confidence interval. Table
III.2 displays the 95% confidence interval for the density estimates of the spring and
summer. This output indicates the accuracy of the population estimates. The traditional
method does not provide a confidence interval.
Distance sampling assumes that all animals are detected from a certain distance.
It adjusts for the probability that individuals at farther distances will not be seen as
readily as individuals close to the observation point. In the trapping web adjustment to
28


distance sampling used in this study, this assumption is translated to the density of the
traps within the trapping web. There is a very high density of traps at the center of the
web and that density decreases farther away from the center. This assumes that all
animals present at the center have a high probability of capture, i.e. 1. The probability of
capture decreases with the decreased density towards the outside of the web. This
mimics the traditional distance sampling, as the observer will have a lower probability of
detecting an animal farther away from the transect line or point. It also does not require
an estimate of trapping area, a measurement that is based off assumptions regarding
animal movement in and out of the trapping grid (Corn and Conroy, 1998).
Perognathus fasciatus has been trapped in other studies conducted in Colorado in
the past decade. The Colorado Natural Heritage Program (CNHP) surveys and monitors
rare and threatened species in Colorado and works to promote biodiversity conservation.
Small mammal surveys have produced several captures of P. fasciatus in Colorado. A
concentrated effort in the summer of 2009 at Maxwell Ranch, run by Colorado State
University Research Foundation (CSURF), in Larimer county, returned a total of 22
captures in an area approximately 1,700 ha and several density estimations. Researchers
used program DENSITY to estimate population densities of P. fasciatus successfully at
three transects and returned estimates of 2.23, 0.55, and 0.55 individuals/ha (J. Siemers,
CNHP, personal communication). These estimates provide additional insight to Colorado
OBPM population densities.
Grassland and semi-desert rodent species density and resulting abundance can
have a large impact on the vegetative community by herbivorous grazing, seed caching,
and seed predation. The species composition and relative abundance of these animals
29


play a role in the degree of the affects on each of these factors (Jones and Longland,
1999). Burrowing animals like the OB PM, also play a unique role in soil aeration,
nutrient cycling, and habitat transformation. These effects can cascade to other tropic
levels and influence overall biodiversity of the ecosystem from bacteria to invertebrates
to carnivores (Davidson et al., 2012).
Home Range
The estimated home ranges from P.fasciatus individuals are the first calculated
for this species. Our estimates are comparable in size to other Perognathus species
(Table III.5). The plains pocket mouse (Perognathus flavescens), a sympatric species
with an overlapping geographic range, has a smaller estimated home range than P.
fasciatus. Perognathus flavescens has been described as more particular to very sandy
soils whereas P. fasciatus is more tolerant of a variety of denser vegetation habitats
(Williams and Genoways 1979). The specialization of P. flavescens might account for the
difference in home range area estimations. Despite the challenges of transmitter
attachment and small sample size, our estimates provide a much-needed increase in our
understanding of the spatial scale required for conservation of the species.
Our results suggest that male home range sizes were larger on average than those
for females. Home range overlap can provide information regarding competition, mating
patterns, and territoriality (Powell, 2000; Burt, 1943). In the spring, the two males
tracked showed 10.87% overlap of their core (50% kernel) areas. In the summer there
was only a small percentage of overlap between 1 male and 1 female at the 50% kernel,
with 1.6% (Figure IV. 1). Within the core 50% kernels, there was no overlap of female-
30


to-female home range areas. This is a common pattern in polygynous species, where
males overlap several female home ranges and mate with multiple females.
Investigation into seasonal variation of home range was limited by small sample
size. Species of Perognathus genus are united in the presence of fur-lined cheek pouches
that allow these animals to gather many seeds in a manner that does not sacrifice
moisture and provides easy transport to burrows for storing food for winter (Vander Wall
et al., 1998). The exact method of hoarding for P. fasciatus is unknown, but other small
pocket mice are primarily larder hoarders, depositing large number of seeds in one place,
like a chamber of a burrow system (Price et al., 2000). If the OBPM uses this method of
storage, it is less likely that the animal would need to expand its home range area in the
late summer and fall as resource availability decreases, preferring to remain close and
provide protection to the burrow that contains the majority of its seed stores.
Ground Cover
The ground cover within the 50% kernel home range area showed less bare
ground and more litter than the literature suggests for the preferred habitat of this species.
Most accounts indicate that the OBPM is found on sandy soils, with sparse vegetation.
Because there are limited studies on P. fasciatus density, it is unclear whether there is a
relationship between ground cover type and population density.
The placement of the trapping web may provide some insight into the relationship
of OBPM population presence and shrub cover. The trapping web covered areas of dense
shrub cover and areas of no shrub cover. Almost all of the OBPM captures occurred on
the half of the web that was free of shrubs (Figure IV. 1). This observation as well as the
almost complete absence of shrub cover in the estimated home ranges suggests that this
31


species prefers land free of shrub cover, possibly relying on grass and forb cover for
protection from predators.
Grasslands with increased litter are a result of less grazing and infrequent fires,
and of increased plant productivity from increased precipitation, which can result in a
decrease of granivores (Reed et al., 2006a). Small mammal species diversity studies
theorize that pocket mice as a group are more often found in habitat with increased bare
ground from less vegetation and reduced litter (Thompson and Gese, 2013). A proposed
explanation for this is the difficulty of locating seeds in excessive litter cover (Reed et al.,
2006b). Our study site had more litter than bare ground, yet still supported a population
of P. fasciatus.
Reitrodontomys megalotis was the most abundant species captured in our study.
This species prefers productive grassland habitat with increased litter (Kaufman et al.,
2000). The dynamics between R megalotis and P. fasciatus are not clear. It has been
suggested that murid population density increases with increasing vegetation, but pocket
mice density increases with decreasing vegetation and litter (Jones et al., 2003).
Temporal effects might also play a part in changing abundances. Other studies support
the trend we observed of low density of R. megalotis in the spring and summer and
higher density in the fall and winter (Table IV.II) (Sullivan and Sullivan, 2008). The
changes in values might reflect the available forage throughout the year, with a summer
abundance deterring entrapment and a winter scarcity encouraging it.
Dominant Plant Species
The dominant vegetation of an area is important in various ways for the resident
small mammal community. The shape and structure of growth can provide protection
32


and nesting material and the herbaceous and reproductive parts are vital food sources for
granivorous and herbivorous rodents. The dominant plant species in an area function as
both food and shelter. The five most dominant plants recorded from the Daubenmire
surveys are all perennials except for Bromus sp. and H. canescens. According to the US
Forest Service, S. cryptandrus and Bromus sp. were rated good for small mammal
consumption, A. purpurea was rated poor, and P. smithii and H. canescens were not
rated (Dittberner and Olson, 1983). Bromus sp. was the only overlapping species found
in this study and previous studies of P.fasciatus habitat, but perennial grasses of various
species were consistent across several studies (Manning and Jones, 1988).
This species is geographically wide-ranging and previous researchers have
reported captures in a variety of habitat type and vegetation (Manning and Jones, 1988).
Our findings support the hypothesis that perennial grasses are dominant species in areas
occupied by multiple OBPMice. However, grasslands of Colorado are generally
composed of a diverse gradient of perennial grasses. This study provides another
description of OBPM habitat, but cannot characterize a quantitative definition of
dominant vegetation common to the OBPM populations. We did not directly consider
soil composition in this study. Soil type, not vegetation, might be the unifying factor of
this animals actual area of occurrence. The home range area of most of the individuals
tracked at the PCC was in a sandy wash.
Suitable Soils for P. fasciatus in Colorado
The role that soil composition plays in the success of P.fasciatus populations is
not well understood. Burrow dwelling animals do require certain foundational needs to
support subterranean living, but also need to be able to excavate soil with relative ease.
33


In a preliminary exploration of soil data, we explored the soils types where individual
OBPM have been captured across Colorado. We retrieved capture data from VertNet
database, a collection of hundreds of university and museum records. These data, in
addition to capture locations collected by CNHP and this study, were combined in
ARCMaplO. 1 with GIS soil data from the USDAs Natural Resource Conservation
Service (NRCS), U.S. General Soil survey. We identified the types of soil where all the
individuals were captured and then selected those various soil types at a state level to
identify suitable soil for OBPM (Figure IV.III). These areas provide some additional
insight for future OBPM trapping studies.
Rare Status Assessment
Setting minimum density thresholds to establish viable populations or degree of
rarity in mammals is very species dependent. Yu and Dobson (2000) established
guidelines for rarity in mammals and proposed the median for small mammals (<100g) at
100 individuals/km2 or 1 individual /ha. Our study provides estimates that are above this
proposed median threshold. When compared to other members of the Perognathus genus
this median measurement is low. Perognathus parvus has been found at densities
between 42 82 individuals/ha in sagebrush habitat at the northern part of their range
(Sullivan and Sullivan, 2008). In the absence of additional density estimate studies, we
can only contribute our estimates from distance sampling to the collection of literature on
this species.
A unique challenge in trapping this particular species is the animals tendency to
go into torpor during extreme temperatures. In addition to the dormancy in the winter
months, P. fasciatus can also enter estivation in very hot temperatures. The summer of
34


2012 set record temperatures in July (http://www.esrl.noaa.gov/psd/boulder/boulder.
davsgt90.htm0. Additional trapping was conducted at Green Mountain in Lakewood,
Colorado between seasonal efforts at the PCC. During the multiple weeks of +90 degree
days, rodent captures were scarce. Once the temperatures returned to average,
Heteromyids were captured (CPW, unpublished).
Colorado Parks and Wildlife first began a concentrated trapping effort for the
OBPM in 2008 when it detected a small population at Woodhouse State Wildlife Area in
Jefferson County, Colorado. Further efforts in 2010 resulted in the capture of only one
individual from a multi week effort at Green Mountain and no individuals at Rocky Flats
National Wildlife Refuge, both in Jefferson County Colorado. The PCC is the only area
where a population was detected in 2010. The scope of these trapping efforts around the
state in likely OBPM habitat supports the idea that these animals are regionally rare,
possibly restricted by habitat or soil type. However, assumptions regarding populations
being locally abundant but regionally scarce would be premature without additional
studies.
The final concept of rarity presented by Rabinowitz (1981) is rare due to
geographic range restriction. The entirety of the OBPM geographic range is beyond the
scope of this study, but within Colorado, there is support for a rapid decrease in the
available habitat. The CGAP analysis report created by CNHP indicates that there is less
than 1% of all grassland type (short, mid and tall) protected under status 1 or 2. These
status rankings indicate land that has full protection from anthropogenic interference and
management that mimics natural processes (status 1), or management that may interfere
with completely natural processes (status 2) (Colorado Gap Analysis Program, 2000).
35


This limited amount of protected area contradicts the IUCN interpretation that this mouse
does occur in protected areas, at least within the state of Colorado. Regardless of
protected lands in states that occupy the northern portion of this animals range, the lack
of secure habitat in the southern periphery of the geographic range can result in local
extinctions that may reduce overall genetic variability to the species. Conservation effort
on peripheral populations is a debated topic, with arguments both for and against (Garner
et al. 2004). Land Managers and Biologists in Colorado should consider the species
overall geographic range, but make decisions at a local level to protect the states
biodiversity. The other factor that supports the IUCN description of the mouse as least
concern is the explanation that there are no major threats. The rapid human population
growth within Colorado is a threat to habitat via conversion and degradation in this
portion of the mouses range.
Conservation of Rare Species in Diminishing Habitat
Conservation of rare species is especially important in areas of rapid habitat loss
to conversion or degradation from human activities. The lesson learned from the PMJM
was to never ignore the importance of a thorough knowledge of species life history,
genetics, demography, home range, and geographic distribution. The information
presented in this study provides novel quantitative values that will aid in the management
of grassland biodiversity.
36


Tables and Figures
Figure IV. 1: Aerial View of the study site with the trapping web outline (A). The pink
polygon is general area of pocket mouse captures in the spring (B) and the green polygon
is general area of summer OBPM captures (C). The right side of the web has a higher
density of shrub cover than the left side where the OBPM were captured.
37


Spring Summer
50% Kernels 50% Kernels
| | D (male) * * A (female)
W (male) V [ j B (male)
/ \ E (female)
I | F (female)
Figure IV.II: The 50% kernels are isolated for each animal and displayed by season with
the spring males showing more overlap (left), than the individuals in summer (right).
38


Sedgwick
Phillips
Cheyenne
40 80
160
HMiles
Figure IV.III: The shaded areas represent all soil types in the state where OB PM have
been captured or collected. The triangles are capture and collection sites compiled from
this study, CNHP records, and VertNet database.
Table IV. 1: R. megcilotis density estimations by season as calculated by program
Distance with a negative exponential function with cosine adjustment model selected by
AIC.
Season D %CV df 95% Confidence interval
Spring 16.0 31.1 35.3 8.7 29.7
Summer 3.8 45.2 42.5 1.6 8.9
Fall 31.4 56.7 7.58 9.1 107.4
Winter 29.8 115.16 5.3 2.9 303.7
39


REFERENCES
ANDERSON, D. R., K. P. BURNHAM, G. C. WHITE, AND D. L. OTIS. 1983. Density
Estimation of Small-Mammal Populations Using a Trapping Web and Distance
Sampling Methods, Ecology 64:674-680.
ANIMAL CARE AND USE COMMITTEE. 1998. Guidelines for the Capture, Handling,
and Care of Mammals as Approved by the American Society of Mammalogists,
Journal of Mammalogy 79:1416-1431.
BEST, T. L., AND M. P. SKUPSKI. 1994. Perognathus flavus, Mammalian Species
471:1-10.
BURNHAM, K. P. AND D. R. ANDERSON. 2002. Model selection and multimodel
inference: a practical information-theoretic approach, 2nd edition, Springer-
Verlag, New York, NY.
BURT, W. H. 1943. Territoriality and Home Range as applied to Mammals, Pp. 346-352,
Journal of Mammalogy.
CORN, J. L., AND M. J. CONROY. 1998. Estimation of density of mongooses with
capture-recapture and distance sampling, Journal of Mammalogy 79:1009-1015.
CRIFASI, R. R. 2007. A subspecies no more? A mouse, its unstable taxonomy, and
western riparian resource conflict, Cultural Geographies 14:511-535.
DAVIDSON, A. D., J. K. DETLING, AND J. H. BROWN. 2012. Ecological roles and
conservation challenges of social, burrowing, herbivorous mammals in the world's
grasslands, Frontiers in Ecology and the Environment 10:477-486.
DITTBERNER, P.L. AND M R.OLSON. 1983. The plant information network (PIN) data
base: Colorado, Montana, North Dakota, Utah, and Wyoming. FWS/OBS-83/86.
Washington, DC: U.S. Department of the Interior, Fish and Wildlife Service. 786,
p.806.
DREVER, C. R., M. C. DREVER, AND D. J. H. SLEEP. 2012. Understanding rarity: A
review of recent conceptual advances and implications for conservation of rare
species, Forestry Chronicle 88:165-175.
FIEBERG, J., AND C. O. KOCHANNY. 2005. Quantifying home-range overlap: The
importance of the utilization distribution, Journal of Wildlife Management
69:1346-1359.
GARNER, T. W. J., P. B. PEARMAN, AND S. ANGELONE. 2004. Genetic diversity
across a vertebrate species' range: a test of the central-peripheral hypothesis,
Molecular Ecology 13:1047-1053.
40


GASTON, K. J. 1991. How large is a species geographic range? Oikos 61:434-438.
GASTON, K. J., AND R. A. FULLER. 2009. The sizes of species' geographic ranges,
Journal of Applied Ecology 46:1-9.
GENOWAYS, H.H. AND J.K. JONES JR. 1972. Mammals from southwestern North
Dakota. Occasional Papers Museum, Texas Tech University, 6:1-36.
HAYDEN, P. and R. LINDBERG. 1976. Survival of laboratory-reared pocket mice,
Perognathus longimembris. J. Mammalogy 57(2):266-272.
HEDLUND, J. D., AND L. E. ROGERS. 1980. Great basin pocket mice Perognathus
parvus in the vicinity of radioactive waste management areas, Northwest Science
54:153-159.
HUNTER, M. L. AND A. HUTCHINSON. 1994. The virtues and shortcomings of
parochialism-conserving species that are locally rare, but globally common,
Conservation Biology 8(4): 1163-1165.
JONES, A. L., AND W. S. LONGLAND. 1999. Effects of cattle grazing on salt desert
rodent communities, American Midland Naturalist 141:1-11.
JONES, Z. F., C. E. BOCK, AND J. H. BOCK. 2003. Rodent communities in a grazed
and ungrazed Arizona grassland, and a model of habitat relationships among
rodents in southwestern grass/shrublands, American Midland Naturalist 149:384-
394.
KAUFMAN, D. W., G. A. KAUFMAN, AND B. K. CLARK. 2000. Small mammals in
native and anthropogenic habitats in the Lake Wilson area of north-central
Kansas, Southwestern Naturalist 45:45-60.
KING, T. L., ET AL. 2006. Comprehensive genetic analyses reveal evolutionary
distinction of a mouse (Zapus hudsoniuspreblei) proposed for delisting from the
US Endangered Species Act, Molecular Ecology 15:4331-4359.
KOTLER, B. P., AND J. S. BROWN. 1988. Environmental heterogeneity and the
coexistence of desert rodents, Annual Review of Ecology and Systematics
19:281-307.
MACKENZIE, D. I., J. D. NICHOLS, N. SUTTON, K. KAWANISHI, AND L. L.
BAILEY. 2005. Improving inferences in popoulation studies of rare species that
are detected imperfectly, Ecology 86:1101-1113.
41


MALANEY, J. L., AND J. A. COOK. 2013. Using biogeographical history to inform
conservation: the case of Preble's meadow jumping mouse, Molecular Ecology
22:6000-6017.
MANNING, R. W., AND J. K. JONES, JR. 1988. Perognathusfasciatus, Mammalian
Species 303:1-4.
MERRITT, J. F. 2010. The biology of small mammals, The Johns Hopkins University
Press, Balitmore, Maryland, USA.
MONK, R. R., AND J. K. JONES, JR. 1996. Perognathus flavescens, Mammalian
Species 525:1-4.
NEISWENTER, S. A., AND B. R. RDDLE. 2011. Landscape and climatic effects on the
evolutionary diversification of the Perognathus fasciatus species group, Journal
of Mammalogy 92:982-993.
ORLAND, M. C., AND D. A. KELT. 2007. Responses of a Heteromyid Rodent
Community to Large- and Small-Scale Resource Pulses: Diversity, Abundance,
and Home-Range Dynamics, Journal of Mammalogy 88:1280-1287.
PARMENTER, R. R., ET AL. 2003. Small-mammal density estimation: A field
comparison of grid-based vs. web-based density estimators, Ecological
Monographs 73:1-26.
PEFAUR, J. E., AND R. S. HOFFMAN. 1974. Notes of the biology of the olive-backed
pocket mouse Perognathus fasciatus on the Northern Great Plains, Prairie
Naturalist 6(1):7-15.
POWELL, R. 200. Animal Home Ranges and Territories and Home Range Estimators,
Pp. 65-110 in Research Techniques in Animal Ecology (L. Boitoni and T. K.
Fuller, eds.). Columbia University Press, Methods and Cases in Conservation
Science.
PRICE, M. V., N. M. WASER, AND S. MCDONALD. 2000. Seed caching by
heteromyid rodents from two communities: Implications for coexistence, Journal
of Mammalogy 81:97-106.
QUIRICI, V., ET AL. 2010. Seasonal variation in the range areas of the diurnal rodent
Octodon degus, Journal of Mammalogy 91:458-466.
RABINOWITZ, D. 1981. Seven forms of rarity, Pp. 205-217 in The biological aspects of
rare plant conservation (H. Synge, ed.). Wiley, Chichester, England.
RAMEY, R. R., H. P. LIU, C. W. EPPS, L. M. CARPENTER, AND J. D. WEHAUSEN.
2005. Genetic relatedness of the Preble's meadow jumping mouse (Zapus
42


hudsoniuspreblei) to nearby subspecies of Z. hudsonius as inferred from variation
in cranial morphology, mitochondrial DNA and microsatellite DNA: implications
for taxonomy and conservation, Animal Conservation 8:329-346.
RANDALL, J. A. 1993. Behavioral adaptations of desert rodents (Heteromyidae),
Animal Behaviour 45:263-287.
REED, A. W., G. A. KAUFMAN, AND D. W. KAUFMAN. 2006a. Effect of plant litter
on seed predation in three prairie types, American Midland Naturalist 155:278-
285.
REED, A. W., G. A. KAUFMAN, AND D. W. KAUFMAN. 2006b. Species richness-
productivity relationship for small mammals along a desert-grassland continuum:
Differential responses of functional groups, Journal of Mammalogy 87:777-783.
SAMSON, F., AND F. KNOPF. 1994. Prairie conservation in North America, Bioscience
44:418-421.
SCHRUPP, D.L., W.A. REINERS, T.G. THOMPSON, L.E. OBRIEN, J.A. KINDLER,
M B. WUNDER, J.F., LOWSKY, J.C. BUOY, L. SATCOWITZ, A.L. CADE,
J.D. STARKS, K.L. DRIESE, T.W. OWENS, S.J., RUSSO, andF. DERCHIA.
2000. Colorado Gap Analysis Program: A Geographic Approach to Planning for
Biological Diversity Final Report, USGS Biological Resources Division, Gap
Analysis Program and Colorado Division of Wildlife, Denver, CO
SULLIVAN, T. P., AND D. S. SULLIVAN. 2008. Dynamics of Peripheral Populations
of Great Basin Pocket Mice, Perognathusparvus, and Western Harvest Mice,
Reithrodontomys megalotis, in Southern British Columbia, Canadian Field-
Naturalist 122:345-356.
THOMPSON, C. M., AND E. M. GESE. 2013. Influence of vegetation structure on the
small mammal community in a shortgrass prairie ecosystem, Acta Theriologica
58:55-61.
VANDER WALL, S. B., W. S. LONGLAND, S. PYARE, AND J. A. VEECH. 1998.
Cheek pouch capacities and loading rates of heteromyid rodents, Oecologia
113:21-28.
WILLIAMS, D.F. AND H.H. GENOWAYS. 1979. A systematic review of the olive-
backed pocket mouse, Perognathus fasciatus (Rodentia:Heteromyidae). Annals of
Carnegie Museum. 48:73-102.
WRIGLEY, R. E., J. E. DUBOIS, AND H. W. R. COPLAND. 1991. Distribution and
ecology of 6 rare species of prairie rodents in Manitoba, Canadian Field-
Naturalist 105:1-12.
43


YU, J. P., AND F. S. DOBSON. 2000. Seven forms of rarity in mammals, Journal of
Biogeography 27:131-139.
44


APPENDIX
Table Al. Density of all rodents at the PCC by season. The model that best fit the data
based on the AIC value was the uniform cosine detectability function with the data
truncated at 65 meters, and the data weakly monotonically non-increasing. The closest A
AIC was 34.76 so no other model was considered. Both rodent density and abundance
were highest in the fall and lowest in the summer.
All rodents D %CV df 95% Confidence interval
Spring 16.045 31 35.28 8.6762 29.672
Summer 3.7708 45.2 42.49 1.5802 8.9983
Fall 31.41 56.71 7.58 9.1853 107.41
Winter 29.774 115.16 5.3 2.9188 303.72
N
Spring 28 31 35.28 15 52
Summer 7 45.2 42.49 3 16
Fall 55 56.71 7.58 16 190
Winter 53 115.16 5.3 5 536
Table A2. Program DISTANCE models used to estimate density of all rodents captured
at the PCC. The most parsimonius model has the lowest AIC score.
Model No. of parameters AAIC
Uniform+cosine 7 0
Negative + cosine 5 34.8
Half normal + cosine 6 35.5
Negative + simple polynomial 4 35.9
Table A3. Reproductive status of P.fasciatus in 2012. Reproductive females include
evidence of lactation or nursing, distended abdomen as evidence of pregnancy, or
perforate vagina. Reproductive males show enlarged testes/scrotal sac.________________________
Season Reproductive Non- Reproductive Non-
Females Reproductive Females Males Reproductive Males
Spring 0 5 1 3
Summer 0 9 0 1
45


Table A4. Age of P.fasciatus in 2012. Individuals identified as adult or sub adult based
on mass and molt patterns. No juveniles were captured.______________________________
Season Adult Females Sub Adult Females Adult Males Sub Adult Males
Spring 5 0 4 0
Summer 8 1 1 0
Table A5. Species richness between 2010 and 2012. Comparison of rodent diversity
between August trapping in 2010 and 2012 at the PCC. Trapping areas were within
500m of one another (see Figure A2),__________________________________
Species 8/6/10-8/13/10 8/1/12-8/8/12
C. hispidus X X
D. ordii X
M. ochrogaster X X
N. floridana X
N. mexicana X
O. leucogaster X
P. fasciatus X X
P. maniciriatus X X
R. megalotis X X
R. montanus X X
Total 10 6
Table A6. Comparison of ground cover between 2010 and 2012 trapping efforts. Vegetation surveys did not share the same methodology, nor did the areas spatially overlap entirely.
Ground Cover Type Aug. 2010 Mean May and Aug. Combine 2012 Mean
Bare Ground 15.3% 4.5-21%
Grass and Forbs 74.7% 31-41%
Litter 10.0% 39 58%
46


Table A7. Soil types returned in the OBPM area of occurrence analysis. Parcel count
refers to total parcels of that particular soil type throughout the state and does not
correspond directly to OBPM presence in each parcel.___________________
Parcel
Count Soil Type______________________________________________________
Winona-Travessilla-Schooner-Rock outcrop-Rentsac-Duffymont-
2 Crago
25 Ryan Park-Rock River-Maybell-Grieves-Crestman-Berlake
2 Starman-Rhone-Parachute-Northwater-Irigul
1 Weld-Stoneham-Platner-Olney-Nunn-Ascalon
1 Ulm-Nunn-Englewood
1 Valmont-Nunn-Nederland-Leyden-Kutch-Denver
2 Truckton-Bresser
1 Cushman-Bresser-Ascalon
5 Peyton-Kettle-Brussett
1 Nunn-Heldt-Haplustolls-Ellicott
13 Woodhall-Rogert-Raleigh
6 Rock outcrop-Resort-Boyle
1 Seitz-Peeler-Granile
1 Manzanola-Kim-Fort Collins
4 Silvercliff-Feltonia-Coutis
1 Wix-Redfeather-Coutis
1 Wahatoya-Ring-Noden-Maitland-Bayerton
4______Travessilla-Rock outcrop______________________________________
47


Spring Movement and Home Range
Slimmer Movement and Home Range
Figure Al. 95% kernel polygons for all animals, by season, with observation points
included.
48


Full Text

PAGE 1

AN ECOLOGICAL STUDY OF PEROGNATHUS FASCIATUS by CHELSEA K. BEEBE B.A. University of Colorado, Boulder 2004 ! ! ! ! "!#$%&'&!&()*'##%+!#,!#$% -./(0#1!,2!#$%!34.+(.#%!5/$,,0!,2!#$% 67'8%4&'#1!,2!9,0,4.+,!'7!:.4#'.0!2(02'00*%7# ,2!#$%!4%;('4%*%7#&!2,4!#$%! +%<4%%!,2 =.&#%4!,2!5/'%7/% >',0,<1 ?@AB ! !

PAGE 2

ii C$'&!#$%&'&!2,4!#$%! =.&#%4!,2!5/'%7/%!+%<4%% !)1 9$%0&%.!DE!>%%)% $.&!)%%7!.::4,8%+!2,4!#$% F%:.4#*%7#!,2!G7#%<4.#'8%!>',0,<1 )1 ! ! ! ! ='/$.%0! >E! H(7+%4 I 9$.'4 J.(4%0! =E! K.4#0%1 9$%4'!L,7%& J.7/%!9.4:%7#%4 ! ! ! ":4'0!?M I ?@AB

PAGE 3

iii >%%)%I!9$%0&%. I! D E !N =E5EI! >',0,<1 O "7!P/,0,<'/.0!5 #(+1!,2! !"#$%&'()*+,-'+./'(*+ ! C$%&'&!+'4%/#%+!)1 !Q4,2%&&,4! ='/$.%0!>E!H(7+%4 ! "#$%&"'% C$%!R0'8% S )./T%+!Q,/T%#!=,(&%!N !"#$%&'()*+,-'+./'(*+ O!'&!.7!(7+%4&#(+'%+! &:%/'%&!U$,&%!&,(#$%47!4.7<%! ,//(:'%&!#$%!9,0,4.+,!-4,7#!V.7<%!.7+!%.&#%47! :0.'7&E! !"//,4+'7
PAGE 4

iv C$%!2,4*!.7+!/,7#%7#!,2!#$'&!.)./#!.4%!.::4,8%+E!!G!4%/,**%7+!'#&!:()0'/.#',7E ! "::4,8%+W! ='/$.%0!>E!H(7+%4! ! ! ! ! ! ! ! ! ! ! ! !

PAGE 5

v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

PAGE 6

vi %"#,-!*1!'*)%-)%$ '2"3%-&! ! GE G[CVRF69CGR[ ! >./T <4,(7+! EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEE A ! K,*%!V .7<% EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEE B ! V.4'#1 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EE B S mall Mammals in a C hang ing L andscape: The C ase of the Preble's Meadow Jumping Mous e EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE M 5#(+1!R )Y%/#'8%& EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEE AA ! GGE =PCKRF5 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEE A? ! 5#(+1!5'#%!F %&/4':#',7 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEE A? ! 5#(+1!F %&'<7 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEE A? ! V.+',!C %0%*%#41 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEE AB ! \%<%#.#',7! 5 (48%1 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEE A] ! F%7&'#1!" 7.01& '& EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEE A^ ! K,*%!V.7<%!" 7.01&'& EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE A^ ! C.)0%&!.7+!'<(4%&! EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEE AM ! GGGE VP56JC5! EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE ?? ! F%7&'#1!.7+!" )(7+.7/%!,2! !0 ,-'+./'(*+ EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEE ?? ! K,*%!V.7<%!5 '_% EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEE ?` ! \%<%#.#',7! EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEE ?` ! C.)0%&!.7+!'<(4%&! EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEE ?B ! G\E FG59655GR[ EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEE ?a

PAGE 7

vii F%7&'#1!.7+!" )(7 + .7/% EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEE ?a ! K,*%!V.7<% EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEE `@ ! 34,(7+!9 ,8%4 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEE `A ! F,*'7.7#!Q0.7#!5 :%/'%& EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEE `? ! 5('#.)0%! &,'0& !2,4 !0,-'+./'(*+ !'7!9,0,4.+ EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEE `` ! V.4%!5#.#(&!" &&%&&*%7# EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEE `B ! 9,7&%48.#',7!,2!V.4 %!5:%/'%&!'7!F'*'7'&$'7
PAGE 8

viii LIST OF TABLES %4567 ! GGGEAE!C,#.0!/.:#(4%&!.7+!4%/.:#(4%&! )1!&:%/'%&bbbbbbbbbbbbbbbbbbbb?B GGGE?E!F%7&'#1!.7+!.)(7+.7/%! ,2 ,!"#$%&'()*+,-'+./'(*+ !)1!&%.&,7bbbbbbbbbbE?B GGGE`E!"G9!8.0(%&!2,4!*,+%0!&%0%/#',7!2,4! !"#$%&'()*+,-'+./'(*+ bbbbbbbbbbbEEE?B GGGEBE!K,*%!4.7<%!.4%. &!,2!'7+'8'+( .0!.7'*.0&bbbbbbbbbbbbbbbbbbbbb?d GGGEdE K,*%!4.7<%!%&#'*.#',7&!,2!,#$%4! !"#$%&'()*+ +1 E EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEE ?d ! GGGE]E!=%+'(*!.7+!'7#%4;(.4#'0%!4.7<%!,2!<4,(7+!/,8%4 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEE ?d G\EAE!F%7&'#1!.7+!.)(7+.7/%!,2! 2"/()#$3$&($45+,4"%'6$(/+ EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEE `M "AE! Density of all rodents at the PCC by season EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEE Bd "?E! "G9!8.0(%&!2,4!*,+%0!&%0%/#',7!2,4! .00!4,+%7#&!)1!&%.&,7 EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEE Bd "`E! Reproductive status of P. fasciatus in 2012 ................................ ............................... 45 "BE! Age of P. fasciatus in 2012 ................................ ................................ ........................ 46 "dE Species Richness between 2010 and 2012 ................................ ................................ 46 "]E Comparison of ground cover between 201 0 and 2012 trapping efforts ..................... 46 "^E Soil types returned in the OBPM area of occurrence analysis ................................ ... 47 ! ! ! !

PAGE 9

ix ,8$%!*1!18/9&-$ 1:;<=7 ! GGEAE!V%&%.4/$!&#(+1!.4%. EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEE AM ! GGE?E!C4.::'7 BE!\%<%#.#',7!&.*:0'7.4%!<4,(7+I!<4.&&Z2,4)I!.7+!0'##%4!),X:0,#& EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE EEEE ?^ ! G\EAE!"%4'.0!8'%U!,2!&#(+1!&'#%!U'#$!#4.::'7
PAGE 10

x LIST OF ABBREVIATIONS AIC Akaike information criterion AO Area of occur rence CGAP Colorado Gap Analysis Program CNHP Colorado Natural Heritage Program CSURF Colorado State University Research Foundation EO Extent of occurrence GPS Global positioning system IACUC Institutional Animal Care and Use Committee IUCN Int ernational Union for Conservation of Nature LSCV Least squared cross validation NRCS Natural Resource Conservation Service OBPM Olive backed Pocket Mouse PIT Passive Integrative Receiver PCC Plains Conservation Center PMJM Preble's Meadow Jum ping Mouse USDA United States Department of Agriculture UTM Universal Transverse Mercator ! ! ! ! !

PAGE 11

2 CHAPTER I INTRODUCTION Background The Olive backed Pocket Mouse, OBPM, ( Perognathus fasciatus) is an understudied species that occurs on the short/m ixed grass prairies of the Colorado Front Range. This uncommon species has not been the subject of many demographic studies presumably because of the challenge of finding and capturing enough individuals for a suitable sample size. The geographic distribu tion range for this species extends from southern Canada south through the Dakotas to the eastern plains of southern Colorado (Ma nning and Jones 1988). Historical captures in Colorado have been reported from numerous sites along the Front Range of the Ro cky Mountains in Larimer, Jefferson, Fremont and Custer counties in Colorado The family Heteromyidae includes several species of pocket mouse. Whereas there is limited literature about P. fas c i atus studies on other heteromyids can be used for a better understanding of the genus Perognathus Pocket mice are thought to be primarily granivorous, collecting seeds in their cheek pouches and caching either underground or on the soil surface (Randall, 1993). Perognathus fasciatus eats various grass seeds an d other herbaceous plants (Manning and Jones 1988). Pocket mice tend to become dormant in the colder winter months, retreating to underground burrows and going into torpor, rousing periodically to feed on food stores (Wrigley et al., 1991). After a perio d of dormancy, pocket mice emerge in the early spring to reproduce before the dry months of summer, but can produce additional litters if favorable conditions exist (Randall, 1993; Wrigley et al. 1991 ). Gestation time is approximately four weeks with

PAGE 12

3 2 6 offspring in the average litter (Manning and Jones, 1988). Some heteromyids enter estivation, a state similar to torpor initiated by very hot temperatures ( Merritt, 2010, p 158 ). The survival and life span of P. fasciatus is not known, but Hayden and L indberg ( 1976 ) estimated life span of P. longimembris a similar sized pocket mouse, to be approximately one year in the wild, with some instances of longevity extending up to three years. As residents of desert ecosystems, much attention has been aff orde d to the Heteromyidae as a model of coexistence and sympatric speciation in harsh environments. Randall (1993) synthesized foraging strategies, microhabitat and space use, and general courtship behavior. Neiswenter and Riddle (2001) disc ussed the family evolution throughout the glacial periods of North America that led to the distinction of P. fasciatus as a unique species Whereas there is some general life history information available for P. fasciatus little quantitative information e xists on the de mography of this specie s Population densities have been estimated as 0.4 4.0 individuals per ha at various northern locations within North Dakota and Montana ( Pefaur and Hoffman, 1974 ; Genoways and Jones, 1972 ). No population density study has been condu cted in the eastern plains of Colorado. In populations of the Great Basin Pocket mouse, P. parvus abundance estimates suggest a seasonal abundance peak in late summer, with the lowest numbers in late winter early spring (Jan. Apr.) (Hedlund and Rogers, 1 980). Estimates of small mammal abundance and density are important for understanding population dynamics. Estimating wildlife population sizes is critical for assessing ecosystem health at the community and population levels. In species that are

PAGE 13

4 unders tudied, density can also provide insight into demographic parameters, community interactions and behavior, as well as home range Home Range W. H. Burt (1943) defined home range as the area traversed by the individual in its normal activities of food gat hering, mating and caring for young. This classic definition has under gone some modifications over the years by several groups (Fieberg and Kochanny, 2005). Home range size and selection are affected by resource availability, inter specific and i ntraspeci fic competition. These factors can be influenced by population density and abunda nce (Orland and Kelt, 2007; Kotl er and Brown, 1988). The relationship s between home range size, extent of overlap, and seasonal shifts are complex. H abitat type, weather and seasonality, predators prey dynamics sympatric species, and unique species populations all likely play influential roles (Quirici et al. 2010 ). There is very little literature about P. fasciatus home range size. The greatest distance between trap locat ions for successive captures of individuals include 26.5 meters and 65.7 meters (Manning and Jones, 1988; Pefaur and Hoffman, 1975). Home range estimates from similar sized Perog nathus range from 0.02 0.63 ha (Monk and Jones, 1996 ; Best and Skupski, 199 4; Hedlund and Rogers, 1980). Further research on the home range size of P. fasciatus in the prairies of Colorado is needed because there are no formal estimates of home range size. Rarity Rare species are at a higher risk of extinction than are more commo n species (Drever et al., 2012). Species populations that are characterized by low abundances are more susceptible to demographic stochasticity, genetic inbreeding effects and extreme

PAGE 14

5 climate events from which population recovery is impossible. Rare spec ies also pose a challenge for collecting basic ecological and demographic data. If a species is not of any particular conservation concern, there might not be any interest in concentrated efforts to conduct studies to fill knowledge gaps. As a result, ra re species are often understudied with limited accounts of sighting, collection, and tracking. A comprehensive understanding of all members of an ecosystem can help define phylogen et ic relationships between similar species or reveal mutualistic or competi tiv e interactions that influence the entire food web. The OBPM is not often captured in small mammal studies and when it is, it is not c aptured in large numbers The perceived rarity of this species suggests a need for further investigation. The concept of rarity is deceptively complex, but a basic definition is a species population with low abundance or limited geographic range (D rever et al., 2012; Gaston, 1991 ). Rabinowitz (1981 ) first described types of rareness in three areas: small g eographic range narrow habitat specificity and small/non dominant po pulation size If a species meets any one of these criteria, it could be defined as a "rare species". Variations of these criteria are further used to identify species that are threatened or endangere d. The International Union for Conservation of Nature (IUCN) red list is widely regarded as the authority on threats to existence of species. Threat level is assessed by several factors that are similar to the Rabinowitz (1981) definition. The IUCN exam ines rate of population decline over time (from historic numbers to current population estimates), range area (including extent of occurrence and area of occurrence), population size and fluctuations, and quantitative analysis that considers demographic

PAGE 15

6 pa tterns into the future at established time intervals N $##:WZZUUUE'(/74%+0'&#E,4
PAGE 16

7 Dakota and Stipa, Bouteloua, Carex at another; Yucca glauca and Pinus ponderosa in Nebraska; Artemisia tridentata Atriplex sp ., and Tetrademia sp in Utah; and Carex, Bouteloua, and Agropyron in Montana (Manning and Jones, 1988). These descriptions suggest g eneral prairie landscapes, but because P. fascia tus is not captured in many trapping studies, there are likely overlooked aspect s of the habitat type that are most important for predicting population existence. According to the IUCN, P. fasciatus is "listed as Least Concern because it is widespread, its populations are secure, there are no major threats, and it occurs in many protected areas throughout its range" ( http://www.iucnredlist.org/details/42608/0 ). The literature cited in the IUCN descri ption of the OBPM is primarily from studies published in the 1970's. Another important consideration when estimating rarity is the probability (p) of detecting an animal given that it is present in the area being surveyed. In mark recapture studies t hat assume a closed population, a common equation used to estimate abundance is N=n/p, where N is the population size, n is the observed count of unique individuals, and p is the probability of detecting an animal given that it is available for capture or observation. Probability of detection, p, can be estimated by repeated measures of observation (such as from a mark recapture type of study), or by modeling it as a function of some covariates, such as distance fro m observer, intensity of effort, etc. Ra re species that exist at low population densities or that are local ly abundant but regionally rare are difficult to encounter. The OBPM tendency towards inactivity in periods of extreme temperatures, either by entering torpor or estivation, poses addition al challenges toward population/individual detection. Collecting enough data points for a mark recapture study can therefore be a challenge (Mackenzie et al., 2005). This reduces the precision of

PAGE 17

8 the estimates for p. Distance sampling is an approach that models detection probability as a function of distance from observer and provides a way to adjust for detection based bias in models that are used to estimate population density or abundance. For a small, nocturnal rodent that is not easy to observe, dist ance sampling can be conducted with trapping webs to account for imperfect detection in the estimation of densities in sparse populations when four assumptions are met (Parmenter et al., 2003). The first assumption is that all animals at the web center ar e captured, and that animal movement is constant throughout the web with no net emigration or immigration. The second assumption is that all animal captures are individual events. The third assumption is that individuals are recorded at their original lo cation and finally, that all distances between traps are measured accurately (Anderson et al., 1983). Measurements of population abundance and geographic range allow for quantitative analysis of rarity, and also provide specific insight for the type of rar ity exhibited (Drever et al., 2012; Rabinowitz, 1981). Distinctions can be made between high/low abundance in one region, or high/low abundance across the entire geographic range and thus inform the subsequent conservation strategies in specific areas, or more importantly, specific states or counties. The organization Nature Serve classifies P. fasciatus as "secure", however in Colorado it is classified as "vulnerable" ( http://natureserve.org/getData/in dex.jsp ). This specification to the Colorado population is an example of how jurisdictional boundaries can influence conservation priorities. The designation of states and countries within the species' geographic range can have harmful and long lasting effects on the survival of the species as conservation and management priorities vary across boundary lines Increased threats at the species range ends or in the

PAGE 18

9 center might cause fragmentation leading to reduced gene flow or loss of genetically divers e edge populations (Drever et al., 2012). Edge populations may carry diverse alleles making them more suited to variable conditions that differ from those in the central part of the range. This genetic diversity can provide valuable traits for the entir e population in shifting geographic ranges that result from climate change (Hutchinson and Hunter, 199 4) The populations of P. fasciatus with the jurisdictional boundary of Colorado are a unique subset of the species. Living at the southern edge of t he entire geographic range, this collection of populations might be adapted to the hotter and drier conditions typical of lower latitudes This area of the Great Plains is also experiencing higher human population growth as compared to the more northern s tates of Wyoming, North Dakota, and Montana. Overall, the Great Plains of North America are in a state of decline (Samson and Knopf 1994). A consequence of rapid conversion of a threatened habitat is the loss of rare species. Compounding this problem a re the threats to rare species at t he periphery of their ranges. Past conservation issues have shown that m ore baseline information about the rodent species f ound in Colorado is needed in order to develop informed land use plans that consider both human l and use and habitat needs for the native fauna. S mall Mammals in a Changing L andscape: The C ase of the Preble's Meadow Jumping Mouse In the 1990's, the Preble's Meadow Jumping mouse, PMJM, ( Zapus hu dsonius preble ) became a species of concern within its lim ited Colorado and Wyoming range. This mouse lives in riparian habitat, a habitat type that is relatively rare and at risk from

PAGE 19

10 development and pollution along the Front Range of the Rockies and along the western edge of the Great Plains. Zapus hu dsonius p reblei is a subspecies of Zapus hu dsonius a species with a much larger geographic range that extends east across the northern US and through much of Canada. There were two conflicting research groups tha t debated the validity of a sub species classificatio n, one arguing that Z.h. preblei was a unique sub species and the other arguing that it was not, based off independent molecular genetics studies and taxonomic interpretations (Crifasi, 2007; King et al., 2006; Ramey et al., 2005). Proponents of the uniqu e sub species pushed for quick moving legislation to protect critical habitat, stressing that the risk of extinction without immediate action was high. Legislation was passed and the mouse was added to the endangered species list as a threatened species in 1998 causing a cascade of changes for landowners and developers. There has since been evidence suggesting that the su bs pecies Z. h. preblei is not a unique taxon, but shares gene flow with other subspecies of Z. hudsonius (Malaney and Co ok, 2013). At pr esent, the sub species is recognized and federally protected, but remains controversial. This debate resulted in copious spending, heightened emotions, and disruption of development (Crifasi, 2007) Regardless of attitudes on the Endangered Species Act an d its impacts on conservation and development, there is a lesson that can be related to other small mammals in the Colorado Front Range. Sound science based policy demands current and accurate information on the life histories, ecology, and evolutionary re lationships among species within the jurisdiction administered by policy makers. Had intensive morphometric, ecological, and genetic studies been preformed on Z. h. preblei prior to over development and degradation of its

PAGE 20

11 habitat, more informed decisions regarding listing could have been made in a timely manner. Conducting post hoc analysis to answer a scientific question rather than hypothesis testing, may lead to co ntroversial interpretations. The Olive backed Pocket Mouse also resides in a habitat ty pe that is at risk of increased conversion from human activities. Should a conflict similar to the PMJM's occur, a body of information on this understudied species must be made available to all stakeholders. Quantitative and accurate estimates of vital d emographic parameters and home range area can be added to existing genetic and speciation information and aid policy makers and interest groups in making informed decisions. Study Objectives The objective of this study was to gather basic field data from a known population of P. fasciatus to measure population density, to calculate home range size, and to describe ground cover characteristics within estimated home ranges. This study will contribute to the sparse literature on this species and aid in futur e prairie management plans and conservation decisions.

PAGE 21

12 CHAPTER II METHODS Study Site Description This study was conducted from 2012 2013 at the Plains Conservation Center (PCC) West Bijou Creek site near the boundary between Arapahoe and Elbert cou nties, near Strasburg, Colorado (39¡ 35' 12.12", 104¡ 16' 38.59). This historically private rangeland was purchased by the PCC in 2006 (Figure II.1). The site continues to support occasional cattle and bison grazing. The vegetation consists primarily of native grasses (e.g. Pascopyrum smithii ) and non native grasses ( Panicum virgatum and Sporobolus cryptandrus ), with a sparsely scattered canopy of rabbitbrush ( Chrysothamnus nauseous) This site was selected because previous survey efforts discovered a po pulation of P. fasciatus here. This area of the Great Plains receives about 39 cm. of rain annually and experiences winter low temperatures of 11 ¡ C and summer high temperatures of 32 ¡ C. Study Design To calculate animal density through distance sampling w e deployed a trapping web consisting of 16 lines radiating from a center point (distance or "d"=0). Each line was 75 meters long with 15 Sherman live traps (3x3.5x9"; H. B. Sherman Traps, Inc., Tallahassee, Florida) spaced five meters apart. There were another six traps at d=0 for a total of 246 traps (Figure II.2). We used a measuring tape to ensure accurate five meter spacing from the center and between subsequent traps We used a compass to set the direction of the 16 lines. The center of the web was placed purposefully at a coordinate where a high density of P. fasciatus was located in 2010 (Colorado Parks and Wildlife, unpublished). This placement was not random because the species is rare; we focused

PAGE 22

13 our efforts in an area where a known populat ion existed. Due to limited equipment and personnel, we p laced just one web in the same location for each seasonal effort. We conducted trapping in each of the four recognized seasons, spring (5 17 May 2012), summer (1 10 August 2012), autumn (2 11 Nove mber 2012), and winter (2 12 January 2013). Trapping events consisted of two sets of four consecutive trap nights, with a two night break between. All traps received a small ball of Polyfill for insulation and were baited approximately 1 hour before dusk with moistened birdseed (Kaytee brand wild bird seed). Traps baited with mois tened birdseed captured more Heteromyids than rolled oats and peanut butter in our 2010 trapping effort (CPW, 2010 unpublished). We checked the traps at dawn each morning. Duri ng most of the autumn session and all of the winter session, traps were also checked between 00:00 01:00. At this time all animals were processed, released, and the traps reopened and then checked again at dawn to minimize exposure. Individuals that were recaptured in a single night were immediately r eleased and only the data collected from the first capture were recorded. All captured animals were identified to species prior to being weighed (to the nearest 0.1g). We measured and recorded standard morp hometrics including tail length, left foot length, ear length and noted the reproductive status for each animal. All captured animals were uniquely marked. Individuals of Perognathus sp ., Chaetodipus hispidus, Microtus ochrogaster, and Dipodomys ordii we re marked with a Passive Integrative Receiver (PIT) tag (Biomark HPT 8, 8.4 MM 134.2 kHz FDXB) that was inserted subcutaneously between the shoulder blades. Animals were anesthetized with an isoflurane inhalant prior to insertion; approximately 3mLs of is oflurane was applied to a cotton ball that was placed inside a metal tea ball. The animal and the tea ball were

PAGE 23

14 secured in a sealed Tupperware container with punctured holes, and we visually monitored each animal for signs that the anesthesia had taken ef fect (e.g. until it was unable to maintain an upright posture). Once anesthetized, the animal was removed from the container and we inserted the PIT tag using a hollow tipped needle (Biomark N165 needle for MK165 implanter). Because we had a limited numbe r of PIT tags, individuals of the most abundant species, Reithrodontomys sp ., were marked by a unique combination of toe amputations No more tha n one toe was taken from the front and back, and we did not remove medial fore fingers All animal protocols w ere approved by the University of Colorado Denver IACUC (protocol# 101836) and were in accordance with the guidelines established by the American Society of Mammalogists for the use o f wild mammals in research (199 8). Radio Telemetry Perognathus fasciatus individuals were captured as part of a trapping effort that was independent from the trapping web effort to estimate density and abundance that was described above. We placed traps in areas where individuals had been previously captured for the mark reca pture study, supplied with Polyfill bedding, and baited with moistened birdseed, and opened approximately one hour before dusk. Traps were checked at dawn. Radio transmitters weighing 0.4 g (model BD 2N; Holohil Systems Ltd., Carp, Ontario, Canada ) were a ttached to P. fasciatus using a harness that was constructed of 0.5mm elastic cord (Bead landing TM ) threaded through a 0.3 mm beading crimp tube (Beadalon ) into a figure eight. The transmitters were then attached to the beading crimp tube using Gorilla G lue brand adhesive and the harness looped over the fore limbs

PAGE 24

15 with the transmitters resting on the animals' dorsal side, as one would wear a backpack (Figure II.3). Each animal was placed in a small plastic terrarium to observe movement, ensuring a proper harness fit and full range of motion. All animals were released at the point of capture once fitted with the transmitter backpack. Animals were subsequently relocated at hourly intervals for the 7 9 hours between dawn and dusk. We used a receiver and a yagi antenna to locate the animals and ap proached the animal within five meters. This distance was initially calibrated using a placed transmitter at a five meter distance and observing the signal strength as a measure of gain and volume on the receiver. We marked the spot with a pin flag, recorded the direction of the strongest signal with a compass bearing and then quickly and quietly moved approximately ten meters to a location parallel to the signal and repeated the measurement. We then returned duri ng the day and recorded the point at which the bearings int ersected using a measuring tape and a GPS (Garmin, GPSmap 60 ). We also placed a color coded pin flag that corresponded to the animal at the intersection point and removed the other two triangula tion pin flags. This allowed for a clear visualization of each animals movement locations throughout the site. We also located each animal underground during the day. Radio telemetry was conducted in the spring and summer only as no P. fasciatus were ca ptured in the fall or winter In the spring, between 24 30 May 2012, three individuals were harnessed. We collected data on two and lost one individual to a predator before any data were collected. In the summer, on 17 August 2012, 4 individuals were har nessed but all transmitters fell off animals after 3 4 days of data collection. The procedure was repeated with new transmitters on 11 September 2012, on 4 animals, three

PAGE 25

16 of which were the same animals from the first summer trial, and one new capture. Th e transmitters fell off these animals again after 2 3 days. Vegetation Survey Directly f ollowing each radio telemetry seasonal effort, we collected vegetation data from the area where the animals were relocated to characterize the habitat used We used a 0.5m x 1m Daubenmire frame and estimated percent composition of bare ground (ground that containe d no visible plant material), grasses and forbs, litt er (dead plant material), and woody shrubs. We also recorded the height of the tallest shrub within th e frame and the width of the shrub at the widest point. Where shrubs were present, we also estimated both the percentage coverage of the shrubs and the undergrowth, so total composition exceeded 100% in some plots. These categories were selected because they are large scale indicators that can be quickly assessed during trap site selection. Previous research also indicates that bare ground shrubs, and litter are important characteristics for Heteromyid foraging and locomotion ( Randall, 1993) The ground cover survey area was created around the recorded triangulation points of nightly animal movement We used a visual assessment of the color coded pin flags to select the survey areas. Since there was much clustering of location points, we delineated a r ectangle around obvious c lusters, leaving at least a two meter border between the outer flags and the survey edge. We then divided the rectangle into two meter by two meter squares with a measuring tape, and sampled each square of the large rectangle with the Daubenmire frame (Figure II.4). Each surveyed square received an identifying code Points that did not easily fit into the survey rectangles were tr eated as outliers. Using the outlying triangulation point as the center, we created a three by three

PAGE 26

17 square survey around the point and repeated the Daubenmire sampling procedure. The outer four corners of all rectangle plots were recorded with a GPS. Coordinate points of each Daubenmire frame were not recorded because the error on the GPS was larger th en the size of the sampled square. Using the outer four corners UTM's of the rectangle plots, we assigned GPS coordinates to each sampled square in ARC Map 10.0 with the fishnet tool. The output fishnet grid points were assigned to the corresponding veg etation sample fra me based on the identifying code assigned in the field (Figure II.5). Density Analysis We used Program DISTANCE version 6.0 to calculate population density. We evaluated the fit for a global detection function for all P. fascia t u s and f or each unique season using uniform, half normal, hazard rate, and negative exponential key functions with a cosine, simple polynomial, and hermite polynomial series expansions. The most parsimonious model was selected using Akaike Information Criterion (A IC) values (Burnham and Anderson 2002). We performed the analysis using both the full set of captures and the data truncated, with all captures past 65 meters excluded. The exclusion of encounters from the outermost rings adjusts for movement of individua ls into the web. Home Range Analysis For each animal, we estimated a kernel home range utilization distribution at 95% and 50% probabilities using the Animal Movement extension of ARC View 3.2 (P. N. Hooge and B. Eichenlaub. 1997. Animal movement extensio n to arcview. ver. 1.1. Alaska Science Center Biological Science Office, U.S. Geological Survey, Anchorage, AK, USA. ). These values were used because this is an unofficial standard throughout

PAGE 27

18 home range literature and will provide results that are comp arable between studies. Additionally, 95% excludes the outer 5% of movement points, points that may indicate an uncommon foray from the home range core. The smoothing parameter (H) was calculated automatically using the Least squared cross validation (LS CV) method. Because we were interested in characterizing areas of high use, we used vegetation plots that were sampled within the estimated 50% kernel for ground cover analysis. Using ARC Map 10.1, we extracted the vegetation point data that fell within the boundaries of the 50% kernel with the Extract point tool. The median and interquartile ranges of each percent ground cover category were calculated using Microsoft Excel for each animal.

PAGE 28

19 Tables and Figures Figure II.1: Study site at t he P lains C onservation C enter in Arapahoe County, Colorado.

PAGE 29

20 Figure II.2: Trapping web consists of 16 lines labeled A P that radiate from a center point. Each spoke ha s 15 Sherman Live Traps spaced five meters apart beginning at d=5m a nd ending at d=75 m. There are six traps in the center at d=0m. Figure II.3: Transmitter attached to harness; P. fasciatus wearing completed harness. !" #" $" % &" '" ( )" *" +" ," -" / 0" 1 N S W E 23456" 789:;<="->?9"@:
PAGE 30

21 Figure II.4: Representation of vegetation survey rectangle, composed of 2m x 2m squares, with Daubenmire frames ( 0.5m x 1m rectangles within). The red flags indicate movement observation points of P. fasciatus individuals. Figure II.5: Vegetation sampling grids created in ArcMap using the fishnet tool, with animal observation points overlaid. !"#$% &$% &$% '$% '$%

PAGE 31

22 CHAPTER III RESULT S In spring, we recorded 156 captures of 60 unique individuals of the following species: Perognathus fasciatus, Peromyscus maniculatus (Deer mouse), Reithrodontomys megalotis (Western Harvest mouse), R. montanus (Plains Harvest mouse) Microtus ochrogaster (Prairie vole), Neotoma mexicanus (Mexican woodrat), Chaetodipus hispidus (Hispid pocket mouse), and Dipodomys ordii (Ord s kangaroo rat) In the summer we recorded 79 captures of 25 individuals including P. fasciatus P. maniculatus, R. megalotis, R. mon tanus, M. ochrogaster, and C. hispidus In the autumn we recorded 185 captures of 68 individuals including R. megalotis, R. montanus P. maniculatus and D. ordii. In the winter we recorded 151 captures of 76 individuals including R. megalotis, P. manicul atus, and D. ordii (Table III.1) Density and Abundance of P. fasciatus The density and the abundance of P. fasciatus were similar in the spring and the summer (Table III.2). A uniform key function model with a polynomial expansion was selected as the mo st parsimonious model for both seasons based on the lowest value for AIC among competing models (Table III.3). The data were not truncated and analysis was based on exact measurements. The shape of the function was constrained as weakly monotonically non i ncreasing. The model predicted an estim ated density of 6.9 individuals /ha in the spring (CV=17.7%) and 7.4 individuals/ha in the summer (CV=17.8%). The detection probability was modeled and estimated by stratum, or season, and returned values of 0.308 for the spring and 0.257 for the summer.

PAGE 32

23 Home Range S ize We collected home range data on six P. fasciatus individuals, two in the spring and four in the summer. The home range areas were similar between seasons (Table III.4), ranging between 0.06 0.84 ha, w ith an average of 0.395 ha. There was much overlap of the 95% kernels for the animals in both seasons (Figure III.1), but less so at the 50% estimation. All male home range areas (mean 0.66 ha, N=3) were larger than those for the females (mean 0.13 ha, N =3). Vegetation We found the median of the measured ground cover percent ages for each animals' home range (Figure III.2). We also calculate the interquartile range for each animal (Table III.6). Shrub cover was minimal and was excluded from the analysis. The five most dominant species in both seasons were Sporobolus cryptandrus (Torr.) A. Gray (Sand dropseed), Heterotheca canescens (DC.) Shinners (grey golden aster), Pascopyrum smithii (Rydb.) . Lšve (Western wheatgrass) Aristida purpurea Nutt. (Purple threeawn), and Bromus sp (Cheat grass).

PAGE 33

24 Tables and Figures Table III.1: Summary of unique individuals (I) captured per season and total recaptured (R) animals. Spring Summer Fall Winter Species I R I R I R I R C. hispidus 1 4 1 6 0 0 0 0 D. ordii 2 3 0 0 2 1 1 1 M. ochrogaster 4 2 1 1 0 0 0 0 N. mexicanus 1 0 0 0 0 0 0 0 P. fasciatus 8 26 7 20 0 0 0 0 P. maniculatus 3 5 1 2 4 3 5 5 R. megalotis 41 93 16 41 59 91 67 104 R. montanus 4 4 2 1 2 2 0 0 Table III.2 : D ensity (indivi duals per hectare) and abundance for P. fasciatus in spring and summer Season D %CV df 95% Confidence interval Spring 6.9355 17.8 38.51 4.8516 9.9146 Summer 7.4131 17.7 30.47 5.1925 10.635 N Spring 12 17.8 38.51 9 18 Summer 13 17.7 30.47 9 19 Table III.3: Program DISTANCE models used to estimate density of P. fasciatus The most parsimoniou s model has the lowest AIC score. Model No. of parameters !AIC AIC score Uni form simp le polynomial 3 0 490.81 N eg ative cos ine 4 6.06 496.87 H al f normal simp le polynomial 2 7.27 498.08 N eg ative simp le polynomial 2 7.75 498.56

PAGE 34

25 Table III.4: Home range area by animal as calculated by the animal movement extension in ARC 3.2 at the 95% kernel P.fasciatus Individual Season Sex 95% fix ed kernel (ha) D Spring Male 0.39 W Spring Male 0.75 A Summer Female 0.16 B Summer Male 0.84 E Summer Female 0.06 F Summer Female 0.17 %4567!888>?@ !K,*%!4. 7 <%!.4%.&!,2!,#$%4! !"#$%&'()*+ !&:%/'%&!#$.#!&$,U!&,*%!:,4#',7! ,2!#$%'4!<%,<4.:$'/!4.7<%!'7!9, 0,4.+,!N5#(+'%&!#$.#!*%.&(4%+!$,*%!4.7<%!&'_%!U%4%! 7,#!/,7+(/#%+!,7!9,0,4.+,!:,:(0.#',7&OE Colorado Perognathus sp. Home Range (Hectares) Perognathus fasciatus .06 .84 Perognathus flavescens a .02 .05 Perognathus flavus b .11 .63 Perognathus parvus c 0 7 .37 =,7T!.7+!L,7%&!AMM] ) >%&#!.7+!5T(:&T'!AMMB / K%+0(7+!.7+!V,<%4&!AMa@ Table III.6: Median and interquartile range for all ground cover type by individuals tracked. Results are from ground cover within the 50% kernels. Spring Summer INDIVIDUAL D W A B E F Grass/Forbs Median 35 45 45 35 30 30 Interquartile Range 25 20 15 25 25 20 Litter Median 45 40 45 47. 5 55 60 Interquartile Range 25 20 35 25 22.5 20 Bare ground Median 15 10 5 5 10 0 GROUND COVER Interquartile Range 20 20 10 15 15 3.75

PAGE 35

26 Figure III.1: 95% kernel density polygon for all individuals in both spring and summer, with imagery base map of la ndscape.

PAGE 36

27 Figure III.2: Median ground cover with first and third quartiles including ba re ground, grass/forbs, and litter; by animal for all animals tracked. a b d e f w 0 20 40 60 80 100 Bareground Cover Animal Percent Cover a b d e f w 0 20 40 60 80 Grass/Forb Cover Animal Percent Cover a b d e f w 0 20 40 60 80 100 Litter Cover Animal Percent Cover

PAGE 37

28 CHAPTER IV DISCUSSION Density and Abundance To our knowledge, this study provided the first estimation of P. fasciatus population density in Colorado. Knowledge of this basic population parameter can be used to estimate actual species numbers that provide valuable information for conservation and management strategies involving habitat preserva tion and space allocation. The estimated density was higher in this study than in previous studies. The differences in densities might reflect different methods used for estimation. Our study used distance sampling, whereas previous studies calculated d ensity by dividing individuals caught by total trapping area (Pefaur and Hoffman 1974, Genoways and Jones 1972). When the data from this study were treated in this way, density estimates were lower (spring was 4.53/Ha and summer 3.40/Ha), falling close to the upper end of the previous estimat es. Early estimation methods that did not consider the probability of detection and required estimates of the effective trapping area are biased towards overestimation (Anderson et al ., 1983). Another benefit of d ista nce sampling for the est imation of population density is the calculation of a 95% confidence interval. Table III.2 displays the 95% confidence interval for the density estimates of the spring and summer. This output indicates the accuracy of the populati on estimates. The traditional method does not provide a confidence interval. Distance sampling assumes that all animals are detected from a certain distance. It adjusts for the probability that individuals at farther distances will not be seen as re adily as individuals close to the observation point. In the trapping web adjustment to

PAGE 38

29 distance sampling used in this study, this assumption is translated to the density of the traps within the trapping web. There is a very high density of traps at the c enter of the web and that density decreases farther away from the center. This assumes that all animals present at the center have a high probability of capture i.e. 1 The probability of capture decreases with the decreased density towards the outside of the web. This mimics the traditional distance sampling, as the observer will have a lower probability of detecting an animal farther away from the transect line or point. It also does not require an estimate of trapping area, a measurement that is bas ed off assumptions regarding animal movement in and out of the trapping grid (Corn and Conroy, 1998). Perognathus fasciatus has been trapped in other studies conducted in Colorado in the past decade. The Colorado Natural Heritage Program (CNHP) surveys a nd monitors rare and threatened species in Colorado and works to promote biodiv ersity conservation. Small mammal surveys have produced several captures of P. fasciatus in Colorado. A concentrated effort in the summer of 2009 at Maxwell Ranch, run by Color ado State University Research Foundation (CSURF) in Larimer county, returned a total of 22 captures in an area approximately 1,700 ha and several density estimations. Researchers used program DENSITY to estimate population densities of P. fasciatus succe ssfully at three transects and returned estimates of 2.23, 0.55, and 0.55 individuals/ha (J. Siemers, CNHP, personal communication). These estimates provide additional insight to Colorado OBPM population densities. Grassland and semi desert rodent species density and resulting abundance can have a large impact on the vegetative community by herbivorous grazing, seed caching, and seed predation. The species composition and relative abundance of these animals

PAGE 39

30 play a role in the degree of the affects on each of these facto rs (Jones and Longland, 1999). Burrowing animals like the OBPM, also play a unique role in soil aeration, nutrient cycling, and habitat transformation. These effects can cascade to other tropic levels and influence overall biodiversity of the ecosystem from bacteria to invertebrates to carnivores (Davidson et al., 2012). Home Range The estimated home ranges from P. fasciatus individuals are the first calculated for this species. Our estimates are comparable in size to other Perognathus spe cies (Table III.5). The plains pocket mouse ( Perognathus flave scens ) a sympatric species with an overlapping geographic range, has a smaller estimated home range than P. fasciatus Perognathu s flavescens has been described as more particular to very sand y soils whereas P. fasciatus is more tolerant of a variety of denser vegetation habitats ( Williams and Genoways 1979 ). The specialization of P. flave scens might account for the difference in home range area estimations. Despite the challenges of transmitte r attachment and small sample size, our estimates provide a much needed increase in our understanding of the spatial scale required for conservation of the species. Our results suggest that male home range sizes were larger on average than those for female s. Home range overlap can provide information regarding competition, mating patterns, and territoriality (Powell, 2000; Burt, 1943). In the spring, the two males tracked showed 10.87% overlap of their core (50% kernel) areas. In the summer there was only a small percentage of overlap between 1 male and 1 female at the 50% kernel, with 1.6% (Figure IV.1). Within the core 50% kernels, there was no overlap of female

PAGE 40

31 to female home range areas. This is a common pattern in polygynous species, where males over lap several female home ranges and mate with multiple females. Investigation into seasonal variation of home range was limited by small sample size. Species of Perognathus genus are united in the presence of fur lined cheek pouches that allow these anim als to gather many seeds in a manner that does not sacrifice moisture and provides easy transport to burrows for storing food for winter (Vander Wall et al., 1998). The exact method of hoarding for P. fasciatus is unknown, but other small pocket mice are p rimarily larder hoarders, depositing large number of seeds in one place, like a chamber of a burrow system (Price et al., 2000). If the OBPM uses this method of storage, it is less likely that the animal would need to expand its home range area in the lat e summer and fall as resource availability decreases, preferring to remain close and provide protection to the burrow that contains the majority of its seed stores. Ground Cover The ground cover within the 50% kernel home range area showed less bare ground and more litter than the literature suggests for the preferred habitat of this species. Most accounts indicate that the OBPM is found on sandy soils, with sparse vegetation. Because there are limited studies on P. fasciatus density, it is unclear whethe r there is a relationship between ground cover type and population density. The placement of the trapping web may provide some insight into the relationship of OBPM population presence and shrub cover. The trapping web covered areas of dense shrub cover a nd areas of no shrub cover. Almost all of the OBPM captures occurred on the half of the web that was free of shrubs (Figure IV.1). This observation as well as the almost complete absence of shrub cover in the estimated home ranges suggests that this

PAGE 41

32 spec ies prefers land free of shrub cover, possibly relying on grass and forb cover for protection from predators. Grasslands with increased litter are a result of less grazing and infrequent fires, and of increased plant productivity from increased precipit ation, which can result in a decrease of granivores (Reed et al., 2006 a ). Small mammal species diversity studies theorize that pocket mice as a group are more often found in habitat with increased bare ground from less vegetation and reduced litter (Thomp son and Gese, 2013). A proposed explanation for this is the difficulty of locating seeds in excessive litter cover (Reed et al., 2006 b ). Our study site had more litter than bare ground, yet still supported a population of P. fasciatus Reitrodontomys me galotis was the most abundant species captured in our study. This species prefers productive grassland habitat with increased litter (Kaufman et al., 2000). The dynamics between R. megalotis and P. fasciatus are not clear. It has been suggested that mur id population density increases with increasing vegetation, but pocket mice density increases with decreasing vegetation and litter (Jones et al., 2003). Temporal effects might also play a part in changing abundances. Other studies support the trend we o bserved of low density of R. megalotis in the spring and summer and higher density in the fall and winter (Table IV.I I ) (Sullivan and Sullivan, 2008). The changes in values might reflect the available forage throughout the year, with a summer abundance de terring entrapment and a winter scarcity encouraging it. Dominant Plant Species The dominant vegetation of an area is important in various ways for the resident small mammal community. The shape and structure of growth can provide protection

PAGE 42

33 and nesting material and the herbaceous and reproductive parts are vital food sources for granivorous and herbivorous rodents. The dominant plant species in an area function as both food and shelter. The five most dominant plants recorded from the Daubenmire surveys are all perennials except for Bromus sp and H. canescens According to the US Forest Service, S. cryptandrus and Bromus sp were rated "good" for small mammal consumption, A. purpurea was rated "poor", and P smithii and H. canescens were not rated ( Ditt berner and Olson, 1983 ). Bromus sp was the only overlapping species found in this study and previous studies of P. fasciatus habitat, but perennial grasses of various species were consistent across several studies (Manning and Jones, 1988). This species is geographically wide ranging and previous researchers have reported captures in a variety of habitat type and vegetation (Manning and Jones, 1988). Our findings support the hypothesis that perennial grasses are dominant species in areas occupied by mult iple OBPMice. However, grasslands of Colorado are generally composed of a diverse gradient of perennial grasses. This study provides another description of OBPM habitat, but cannot characterize a quantitative definition of dominant vegetation common to th e OBPM populations. We did not directly consider soil composition in this study. Soil type, not vegetation, might be the unifying factor of this animals' actual area of occurrence. The home range area of most of the individuals tracked at the PCC was in a sandy wash. Suitable Soils for P. fasc i a tus in Colorado The role that soil composition plays in the success of P. fasciatus populations is not well understood. Burrow dwelling animals do require certain foundational needs to support subterranean living but also need to be able to excavate soil with relative ease.

PAGE 43

34 In a preliminary exploration of soil data, we explored the soils types where individual OBPM have been captured across Colorado. We retrieved capture da ta from VertNet database, a collection of hundreds of u niversity and museum records. These data, in addition to capture locations collected by CNHP and this study were combine d in ARCMap10.1 with GIS soil data from the USDA's Natural Resource Conservation Service ( NRCS ) U.S. General Soil su rvey. We identified the type s of soil w h ere all the individuals were captured and then selected those various soil types at a state level to identify "suitable" soil for OBPM (Figure IV.II I ). These areas provide some additional insight for future OBPM tr apping studies. Rare Status Assessment Setting minimum density thresholds to establish viable populations or degree of rarity in mammals is very species dependent. Yu and Dobson (2000) established guidelines for rarity in mammals and proposed the media n for small mammals (<100g) at 100 individuals/km 2 or 1 individual /ha Our study provides estimates that are above th is proposed median threshold. When compared to other members of the Perognathus genus this median measurement is low. Perognathus parvu s has been found at densities between 42 82 individuals/ha in sagebrush habitat at the northern part of their range (Sullivan and Sullivan 2008). In the absence of additional density estimate studies, we can only contribute our estimates from distance s ampling to the collection of literature on this species. A unique challenge in trapping this particular species is the animals tendency to go into torpor during extreme temperatures. In addition to the dormancy in the winter months, P. fasciatus can also enter estivation in very hot temperatures. The summer of

PAGE 44

35 2012 set record temperatures in July ( http://www.esrl.noaa.gov/psd/boulder/boulder daysgt90.html ) Additional trapping was conducted at G reen Mountain in Lakewood, Colorado between seasonal efforts at the PCC. During the multiple weeks of +90 degree days, rodent captures were scarce. Once the temperatures returned to average, Heteromyids were captured (CPW, unpublished) Colorado Parks and Wildlife first began a concentrated trapping effort for the OBPM in 2008 when it detected a small po pulation at Woodhouse State Wildlife A rea in Jefferson County, Colorado Further efforts in 2010 resulted in the capture of only one individual from a multi week effort at Green Mountain and no individuals at Rocky Flats N ational Wildlife Refuge, both in Jefferson County Colorado. The PCC is the only area where a population was detected in 2010 The scope of these trapping efforts around the state in l ikely OBPM habitat supports the idea that these animals are regionally rare, possibly restricted by habitat or soil type. However, assumptions regarding populations being locally abundant but regionally scarce would be premature without additional studies The final concept of rari ty presented by Rabinowitz (1981 ) is rare due to geographic range restriction. The entirety of the OBPM geographic range is beyond the scope of this study, but within Colorado, there is support for a rapid decrease in the availa ble habitat. The CGAP analysis report created by CNHP indicates that there is less than 1% of all grassland type (short, mid and tall) protected under status 1 or 2. These status rankings indicate land that has full protection from anthropogenic interfer ence and management that mimics natural processes (status 1), or management that may interfere with completely natural processes (status 2) (Colorado Gap Analysis Program, 2000).

PAGE 45

36 This limited amount of protected area contradicts the IUCN interpretation th at this mouse does occur in protected areas, at least within the state of Colorado. Regardless of protected lands in states that occupy the northern portion of this animals range, the lack of secure ha bitat in the southern periphery of the geographic rang e can result in local extinctions that may reduce overall genetic variability to the species. Conservation effort on peripheral populations is a debated topic, with argument s both for and agains t (Garner et al. 2004). Land M anagers and Biologists in Col orado should consider the species overall geographic range, but make decisions at a local level to protect the state s biodiversity. The other factor that supports the IUCN description of the mouse as least concern is the explanation that there are no majo r threats. The rapid human population growth within Colorado is a threat to habitat via conversion and degradation in this portion of the mouse's range. Conservation of Rare Species in Diminishing Habitat C onservation of rare species is especially import ant in areas of rapid habitat loss to conversion or degradation from human activities. The lesson learned from the PMJM was to never ignore the importance of a thorough knowledge of species' life history, genetics, demography, home range, and geographic d istribution. The information presented in this study provides novel quantitative values that will aid in the management of grassland biodiversity.

PAGE 46

37 Tables and Figures A. B. C. Figure IV.1: Aerial View of the study sit e with the trapping web outl ine (A). The pink p olygon is general area of pocket mouse captures in the spring (B) and the green polygon is general area of summer OBPM captures (C) The right side of the web has a higher density of shrub cover than the left side where the OBPM were c aptured.

PAGE 47

38 Figure IV.II : The 50% kernels are isolated for each animal and displayed by season with the spring males showing more overlap (left), than the individuals in summer (right).

PAGE 48

39 Figure IV.II I : The shaded areas represent all soil types in the stat e where OBPM have been captured or collected. The triangles are capture and collection sites compiled from this study, CNHP records, and VertNet database. Table IV.1: R. megalotis density estimations by season as calculated by program Distance with a ne gative exponential function with cosine adjustment model &%0%/#%+!)1! "G9E Season D %CV df 95% Confidence interval Spring 16.0 31.1 35.3 8.7 29.7 Summer 3.8 45.2 42.5 1.6 8.9 Fall 31.4 56.7 7.58 9.1 107.4 Winter 29.8 115.16 5.3 2.9 303.7 E ! ! ! ! ! # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # # W e l d M o f f a t M e s a B a c a P a r k R o u t t Y u m a L a s A n i m a s G a r f i e l d L i n c o l n L a r i m e r P u e b l o G u n n i s o n B e n t S a g u a c h e E l b e r t G r a n d L o g a n R i o B l a n c o E a g l e K i o w a E l P a s o M o n t r o s e O t e r o D e l t a W a s h i n g t o n K i t C a r s o n L a P l a t a J a c k s o n P r o w e r s F r e m o n t M o n t e z u m a H u e r f a n o P i t k i n C h e y e n n e A d a m s M o r g a n C o s t i l l a C o n e j o s A r c h u l e t a D o l o r e s C h a f f e e M i n e r a l H i n s d a l e S a n M i g u e l C u s t e r D o u g l a s C r o w l e y P h i l l i p s B o u l d e r A r a p a h o e A l a m o s a L a k e R i o G r a n d e S e d g w i c k T e l l e r O u r a y J e f f e r s o n S u m m i t S a n J u a n C l e a r C r e e k G i l p i n D e n v e r B r o o m f i e l d 0 8 0 1 6 0 4 0 M i l e s

PAGE 49

40 RE FERENCES ANDERSON, D. R., K. P. BURNHAM, G. C. WHITE, AND D. L. OTIS. 1983. Density Estimation of Small Mammal Populations Using a Trapping Web and Distance Sampling Methods, Ecology 64 : 674 680. ANIMAL CARE AND USE COMMITTEE. 1998. Guidelines for the Capture, Handling, and Care of Mammals as Approved by the American Society of Mammalogists, Journal of Mammalogy 79 : 1416 1431. BEST, T. L., AND M. P. SKUPSKI. 1994. Perognathus flavus Mammalian Species 471 : 1 10. BURNHAM, K. P. AND D. R. ANDERSO N. 2002. Model selection and multimodel inference: a practical information theoretic approach, 2 nd edition, Springer Verlag, New York, NY. BURT, W. H. 1943. Territoriality and Home Range as applied to Mammals, Pp. 346 352, Journal of Mammalogy. CORN, J. L., AND M. J. CONROY. 1998. Estimation of density of mongooses with capture recapture and distance sampling, Journal of Mammalogy 79 : 1009 1015. CRIFASI, R. R. 2007. A subspecies no more? A mouse, its unstable taxonomy, and western riparian resource confli ct, Cultural Geographies 14 : 511 535. DAVIDSON, A. D., J. K. DETLING, AND J. H. BROWN. 2012. Ecological roles and conservation challenges of social, burrowing, herbivorous mammals in the world's grasslands, Frontiers in Ecology and the Environment 10 : 477 4 86. DITTBERNER, P.L. AND M.R.OLSON. 198 3. The plant information network (PIN) data base: Colorado, Montana, North Dakota, Utah, and Wyoming. FWS/OBS 83/86. Washington, DC: U.S. Dep artment of the Interior, Fish and Wildlife Service. 786, p.806. DREVER, C. R., M. C. DREVER, AND D. J. H. SLEEP. 2012. Understanding rarity: A review of recent conceptual advances and implications for conservation of rare species, Forestry Chronicle 88 : 165 175. FIEBERG, J., AND C. O. KOCHANNY. 2005. Quantifying home range overla p: The importance of the utilization distribution, Journal of Wildlife Management 69 : 1346 1359. GARNER, T. W. J., P. B. PEARMAN, AND S. ANGELONE. 2004. Genetic diversity across a vertebrate species' range: a test of the central peripheral hypothesis, Mole cular Ecology 13 : 1047 1053.

PAGE 50

41 GASTON, K. J. 1991. How large is a species geographic range ? Oikos 61 : 434 438. GASTON, K. J., AND R. A. FULLER. 2009. The sizes of species' geographic ranges, Journal of Applied Ecology 46 : 1 9. GENOWAYS, H.H. AND J.K. JONES J R. 1972. Mammals from southwestern North Dakota. Occasional Papers Museum, Texas Tech University, 6:1 36. HAYDEN, P. and R. LINDBERG 1976. Survival of laboratory reared pocket mice, Perognathus longimembris J. Mammalogy 57(2):266 272. HEDLUND, J. D., AND L. E. ROGERS. 1980. Grea t basin pocket mice Perognathus parvus in the vicinity of radioactive waste management areas Northwest Science 54 : 153 159. HUNTER, M. L. AND A. HUTCHINSON. 1994. The virtues and shortcomings of parochialism conserving species that are locally rare, but globally common, Conservation Biology 8(4):1163 1165. JONES, A. L., AND W. S. LONGLAND. 1999. Effects of cattle grazing on salt desert rodent communities, American Midland Naturalist 141 : 1 11. JONES, Z. F., C. E. BOCK, AND J. H BOCK. 2003. Rodent communities in a grazed and ungrazed Arizona grassland, and a model of habitat relationships among rodents in southwestern grass/shrublands, American Midland Naturalist 149 : 384 394. KAUFMAN, D. W., G. A. KAUFMAN, AND B. K. CLARK. 2000 Small mammals in native and anthropogenic habitats in the Lake Wilson area of north central Kansas, Southwestern Naturalist 45 : 45 60. KING, T. L., ET AL. 2006. Comprehensive genetic analyses reveal evolutionary distinction of a mouse ( Zapus hudsonius pr eblei ) proposed for delisting from the US Endangered Species Act, Molecular Ecology 15 : 4331 4359. KOTLER, B. P., AND J. S. BROWN. 1988. Environmental heterogeneity and the coexistence of desert rodents Annual Review of Ecology and Systematics 19 : 281 307. MACKENZIE, D. I., J. D. NICHOLS, N. SUTTON, K. KAWANISHI, AND L. L. BAILEY. 2005. Improving inferences in popoulation studies of rare species that are detected imperfectly, Ecology 86 : 1101 1113.

PAGE 51

42 MALANEY, J. L., AND J. A. COOK. 2013. Using biogeographica l history to inform conservation: the case of Preble's meadow jumping mouse, Molecular Ecology 22 : 6000 6017. MANNING, R. W., AND J. K. JONES, JR. 1988. Perognathus fasciatus Mammalian Species 303 : 1 4. MERRITT, J. F. 2010. The biology of small mammals, T he Johns Hopkins University Press, Balitmore, Maryland, USA. MONK, R. R., AND J. K. JONES, JR. 1996. Perognathus flavescens Mammalian Species 525 : 1 4. NEISWENTER, S. A., AND B. R. RIDDLE. 2011. Landscape and climatic effects on the evolutionary diversif ication of the Perognathus fasciatus species group, Journal of Mammalogy 92 : 982 993. ORLAND, M. C., AND D. A. KELT. 2007. Responses of a Heteromyid Rodent Community to Large and Small Scale Resource Pulses: Diversity, Abundance, and Home Range Dynamics, Journal of Mammalogy 88 : 1280 1287. PARMENTER, R. R., ET AL. 2003. Small mammal density estimation: A field comparison of grid based vs. web based density estimators, Ecological Monographs 73 : 1 26. PEFAUR, J. E., AND R. S. HOFFMAN. 1974 Notes of the biol ogy of the olive backed pocket mouse Perognathus fasciatus on the Northern Great Plains, Prairie Naturalist 6(1):7 15 POWELL, R. 200. Animal Home Ranges and Territories and Home Range Estimators, Pp. 65 110 in Research Techniques in Animal Ecology (L. Bo itoni and T. K. Fuller, eds.). Columbia University Press, Methods and Cases in Conservation Science. PRICE, M. V., N. M. WASER, AND S. MCDONALD. 2000. Seed caching by heteromyid rodents from two communities: Implications for coexistence, Journal of Mammal ogy 81 : 97 106. QUIRICI, V., ET AL. 2010. Seasonal variation in the range areas of the diurnal rodent Octodon degus Journal of Mammalogy 91 : 458 466. RABINOWITZ, D. 1981. Seven forms of rarity, Pp. 205 217 in The biological aspects of rare plant conservat ion (H. Synge, ed.). Wiley, Chichester, England. RAMEY, R. R., H. P. LIU, C. W. EPPS, L. M. CARPENTER, AND J. D. WEHAUSEN. 2005. Genetic relatedness of the Preble's meadow jumping mouse ( Zapus

PAGE 52

43 hudsonius preblei ) to nearby subspecies of Z. hudsonius as inf erred from variation in cranial morphology, mitochondrial DNA and microsatellite DNA: implications for taxonomy and conservation, Animal Conservation 8 : 329 346. RANDALL, J. A. 1993. Behavioral adaptations of desert rodents (Heteromyidae), Animal Behaviour 45 : 263 287. REED, A. W., G. A. KAUFMAN, AND D. W. KAUFMAN. 2006a. Effect of plant litter on seed predation in three prairie types, American Midland Naturalist 155 : 278 285. REED, A. W., G. A. KAUFMAN, AND D. W. KAUFMAN. 2006b. Species richness productivi ty relationship for small mammals along a desert grassland continuum: Differential responses of functional groups, Journal of Mammalogy 87 : 777 783. SAMSON, F., AND F. KNOPF. 1994. Prairie conservation in North America Bioscience 44 : 418 421. SCHRUPP, D.L ., W.A. REINERS, T.G. THOMPSON, L.E. O'BRIEN, J.A. KINDLER, M.B. WUNDER, J.F., LOWSKY, J.C. BUOY, L. SATCOWITZ, A.L. CADE, J.D. STARKS, K.L. DRIESE, T.W. OWENS, S.J., RUSSO, and F. D'ERCHIA. 2000. Colorado Gap Analysis Program: A Geographic Approach to Pla nning for Biological Diversity Final Report, USGS Biological Resources Division, Gap Analysis Program and Colorado Division of Wildlife, Denver, CO SULLIVAN, T. P., AND D. S. SULLIVAN. 2008. Dynamics of Peripheral Populations of Great Basin Pocket Mice, Perognathus parvus and Western Harvest Mice, Reithrodontomys megalotis in Southern British Columbia, Canadian Field Naturalist 122 : 345 356. THOMPSON, C. M., AND E. M. GESE. 2013. Influence of vegetation structure on the small mammal community in a shor tgrass prairie ecosystem, Acta Theriologica 58 : 55 61. VANDER WALL, S. B., W. S. LONGLAND, S. PYARE, AND J. A. VEECH. 1998. Cheek pouch capacities and loading rates of heteromyid rodents, Oecologia 113 : 21 28. WILLIAMS, D.F. AND H.H. GENOWAYS. 1979. A syst ematic review of the olive backed pocket mouse, Perognathus fasciatus (Rodentia:Heteromyidae). Annals of Carnegie Museum. 48:73 102. WRIGLEY, R. E., J. E. DUBOIS, AND H. W. R. COPLAND. 1991. Distribution and ecology of 6 rare species of prairie rodents in Manitoba Canadian Field Naturalist 105 : 1 12.

PAGE 53

44 YU, J. P., AND F. S. DOBSON. 2000. Seven forms of rarity in mammals, Journal of Biogeography 27 : 131 139.

PAGE 54

45 APPENDIX Table A1. Density of all rodents at the PCC by season. The model that b es t fit the data based on the AIC value was the uniform cosine detectability function with the data truncated at 65 meters, and the data weakly monotonically non increasing. The closest AIC was 34.76 so no other model was considered. Both rodent density and abundance were highest in the fall and lowest in the summer. All rodents D %CV df 95% Confidence interval Spring 16.045 31 35.28 8.6762 29.672 Summer 3.7708 45.2 42.49 1.5802 8.9983 Fall 31.41 56.71 7.58 9.1853 107.41 Winter 29.774 115.16 5.3 2.9188 303.72 N Spring 28 31 35.28 15 52 Summer 7 45.2 42.49 3 16 Fall 55 56.71 7.58 16 190 Winter 53 115.16 5.3 5 536 Table A2 Program DISTANCE models used to estimate density of all rodents captured at the PCC The most parsimonius model h as the lowest AIC score. Table A3 Reproductive status of P. fasciatus in 2012. Reproductive females include ev idence of lactation or nursing, distended abdomen as evidence of pregnancy, or perforate vagina. Reproductive males show enlarged testes/scrotal sac. Season Reproductive Females Non Reproductive Females Reproductive Males Non Reproductive Males Spring 0 5 1 3 Summer 0 9 0 1 Model No. of parameters !AIC Uniform+cosine 7 0 Negative + cosine 5 34.8 Half normal + cosine 6 35.5 Negative + simple polynomial 4 35.9

PAGE 55

46 Table A4. Age of P. fasciatus in 2012. Individuals identified as adult or sub adult based on mass and molt patterns. No juveniles were captured. Season Adult Females Sub Adult Females Adult Males Sub Adult Males Spring 5 0 4 0 Summer 8 1 1 0 Table A5. Species r ichness between 2010 and 2012. Comparison of rodent diversity between August trapping in 2010 and 2012 at the PCC. Trapping areas were within 500m of one another (see Figure A2). Species 8/6/10 8/13/10 8/1/12 8/8/1 2 C. hispidus X X D. ordii X M. o chrogaster X X N. floridana X N. mexicana X O. leucogaster X P. fasciatus X X P. maniculatus X X R. megalotis X X R. montanus X X Total 10 6 Table A6. Comparison of ground cover between 2010 and 2012 trap ping efforts. Vegetation surveys did not share the same methodology, nor did the areas spatially overlap entirely. Ground Cover Type Aug. 2010 Mean May and Aug. Combine 2012 Mean Bare Ground 15.3% 4.5 21% Grass and Forbs 74.7% 31 41% Litter 10.0% 39 58%

PAGE 56

47 Table A7. Soil types returned in the OBPM area of occurrence analysis. Parcel count refers to total parcels of that particular soil type throughout the state and does not correspond directly to OBPM presence in each parcel. Parcel Count Soil Type 2 Winona Travessilla Schooner Rock outcrop Rentsac Duffymont Crago 25 Ryan Park Rock River Maybell Grieves Crestman Berlake 2 Starman Rhone Parachute Northwater Irigul 1 Weld Stoneham Platner Olney Nunn Ascalon 1 Ulm Nunn Englewood 1 Valmon t Nunn Nederland Leyden Kutch Denver 2 Truckton Bresser 1 Cushman Bresser Ascalon 5 Peyton Kettle Brussett 1 Nunn Heldt Haplustolls Ellicott 13 Woodhall Rogert Raleigh 6 Rock outcrop Resort Boyle 1 Seitz Peeler Granile 1 Manzanola Kim Fort Collins 4 Silvercliff Feltonia Coutis 1 Wix Redfeather Coutis 1 Wahatoya Ring Noden Maitland Bayerton 4 Travessilla Rock outcrop

PAGE 57

48 Figure A1. 95% kernel polygons for all animals, by season, with observation points included.

PAGE 58

49 Figure A2. PCC study site including the 2010 transects, represented by dashed lines, the movement locations of all OBPM tracked in the spring and the summer of 2012, and the outline of the trapping web from 2012 2013, depicted by a circle outline (note that the trapping web i s not precisely to scale). !" #" $" % &" '" ( )" *" +" ," -" / 0" 1 N S W E 23456" 789:;<="->?9"@: